Duval.29. Razonamiento Geometrico

British Society for Research into Learning Mathematics
Proceedings of the Day Conferences held at King's College London, Saturday 28th February 1998 and at the University of Birmingham, Friday and Saturday, 19th and 20th June 1998
These proceedings consist of short papers which were written for the BSRLM day conferences in February and June 1998. The aim of the proceedings is to communicate to the research community the collective research represented at BSRLM conferences, as quickly as possible. We hope that members will use the proceedings to give feedback to the authors and that through discussion and debate we will develop an energetic and critical research community. We particularly welcome presentations and papers from new researchers.
ISSN 1463-6840
Contents
Research Reports presented on February 28th 1998
A Computer-Based Mathematics Laboratory for Undergraduate Chemistry Students: A Preliminary Evaluation Report Phillip Kent, Imperial College, London and Ian Stevenson, Institute of Education, London A Survey of Public Images of Mathematics Urn Chap Sam and Paul Ernest University of Exeter Modelling, Strategic Questioning and the Laws of Student Reasoning in ALevel Mechanics Stuart Rowlands, Ted Graham and John Berry University of Plymouth Investigatory Approaches in Mathematics Teaching - Implications for Teacher Training in India Dr. Vijayalakshmi Chilakamarri, King's College London
Page No.
15
23
Working Group Reports from February 28th 1998

Geometry Working Group Convenor: Keith Jones University of Southampton Advanced Mathematical Thinking Working Group Stephen Hegedus University of Southampton 29
35
Research Reports presented at the New Researchers Day Conference on 19th June 1998
Relations Between Teacher's Representations and Pupil's Images Chris Bills University of Warwick Visualisation and Using Technology in A Level Mathematics Sally Elliott Sheffield Hallam University "What Can We All Say?" Dynamic Geometry in a Whole-Class Zone of Proximal Development John Gardiner Sheffield Hallam University Cabri as a Cognitive Tool Bibi Lins 39
45
51
57
University of Bristol
Primary Children's Imagery in Arithmetic Sandra Pendlington University of Exeter
61
Research Reports presented on June 20th 1998

An Activity Approach to Teaching and Assessment Chris Day South Bank University Mathematical Support for Engineering and Science Students Peter Gill King's College London Tutors' Reflections upon the Difficulties of Learning and Teaching Mathematics at University Level: A Report of Work- in-Progress Elena Nardi University of Oxford Convention or Reality Melissa Rodd, Open University Margaret Barber, Thomas Telford School Mind the 'Gaps': Primary Trainees' Mathematics Subject Knowledge Tim Rowland, Caroline Heal, Patti Barber and Sarah Martyn Institute of Education, London Influences on Student Teachers of Mathematics Jim Smith Sheffield Hallam University Possibilities in Pierce's Existential Graphs for Logic Education Adam Vile and Simon Polovina South Bank University What Can be Learnt by Selecting Anecdotes from a Range of Data? Exemplifying "Noteworthy" Mathematics with a Small Number of Examples Anne Watson University of Oxford Teacher Trainee Students' Understanding of Operation Signs David Womack University of Manchester 67
73
79
85
91
97
103
111
117
Working Group Reports from June 20th 1998

Geometry Working Group Convenor: Keith Jones University of Southampton Semiotics and Mathematics Education Working Group Convenors: Adam Vile, South Bank University Paul Ernest, University of Exeter 123
129
A COMPUTER-BASED MATHEMATICS LABORATORY FOR UNDERGRADUATE CHEMISTRY STUDENTS: A PRELIMINARY EVALUATION REPORT
Phillip Kent, Mathematics Department, Imperial College, London SW7; Ian Stevenson, Mathematical Sciences, Institute of Education, London WCl.
This is a preliminary report from an evaluation of a first-year mathematics course for undergraduate chemists which is designed around a combination of motivatory lectures and assignments in a "mathematics laboratory", where students use the computer algebra package, Mathematica. A student questionnaire was administered and a half-dozen follow-up interviews were done. We report here on the findings from those, using summary data from the sample of 57 students, and a comparison of the interview data from two typical students (one with, and one without, A level mathematics).
Mathematics in the Chemistry course at Imperial College

The "mathematics laboratory" has been under development since 1994, and you can read about its earlier history in (Templer et aI, in press). We have been evaluating its latest version, which includes, for the first time this year, "chemical-mathematical" coursework assignments which require the students to think explicitly about how mathematics is used in chemistry to model chemical processes. The thinking behind this has been described by Ramsden & Templer (in press): We no longer see the relationship between the two subjects as strictly a matter of "learn the maths, then apply it to the science". We are trying to introduce mathematics as a natural, integral part of chemistry; Mathematica's power gives the students the opportunity to do this in an authentic, uncontrived way, with something of the flavour of real research. As one might expect students experience a mixture of emotions when presented with this sort of challenge. One of these is certainly shock. We have deliberately set out to stretch students, but not in a way that they would expect. Their anticipation is that they will learn the techniques and tricks which most perceive as being the proper realm of mathematical experience. Instead we present them with the sort of chemical problems found in research, and with the help of a powerful mathematical toolbox, ask them to investigate the chemistry using those tools.
From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 1
That is not to say that they are not learning any mathematics, far from it. The problems are designed in such a way that they have to understand the mathematical processes that are being performed and the limitations that these may impart on their results. In the evaluation, besides trying to assess pragmatically how well different elements of the course are working, we are trying to get a feel for how students are dealing with the triangle of relationships: MATHEMATICAL KNOWLEDGE SCIENTIFIC KNOWLEDGE
COMPUTER MATHEMATICS SOFTWARE Weare interested, among other things, in the nature and extent of the students' mathematical development, whilst working within a computational environment such as Mathematica. To that end, we present here some initial reflections on the questionnaire data, and interviews that we have conducted.
Backgrounds of the students

One can take it as read that all Imperial College chemistry students are highly qualified in terms of A level grades: 28-30 grade points is usual. What does matter, given the mathematical intensity of the firstyear course, is the combination of subjects studied at A level. Out of the questionnaire sample of 57, 38 had single A level mathematics, 9 had double mathematics, and 10 no A level mathematics. Of those students who did not have A level mathematics, all 10 had Physics (at either an A or B grade), and 7 had Biology as their third A level; their GCSE results ranged from two starred A's to three B' s, and two students also had A-S Mathematics. We were particularly interested in this non-A-Ievel group. They have the double challenge of using mathematics, which they are generally unfamiliar with, in the context of learning chemistry. The significance of having Physics A level is that some mathematics may be familiar from there. It is an open question as to whether students are able to either recognise their physics experience as "mathematics", or "transfer" their knowledge across subject boundaries.
A comparison of two students

To illustrate some of the issues associated with the non-A-Ievel group, we will compare two (male) students' responses to the experience of doing the assignment:
Student J. A levels: Chemistry, Biology, Physics; GCSE Maths. Has used graphical calculators a
little in school, but computers not at all.
Page 2
Student M. A levels: Chemistry, Physics, Maths, AS Further Maths (MEI modular courses). Extensive use of graphical calculator, and some computer use (e.g. BASIC programming). At the stage where we take up the story, the students had just completed a Mathematica-based assignment in which they had to fit curves, based on a quite complex sum-of-exponential-terms model, to a set of genuine experimental data. The interview data concerns general mathematical background, general computing experience (both J and M have their own computers), overall experience with the mathematics course, and work on the chemicalmathematical assignment. We concentrate here on just three issues for purposes of comparison, which are relevant to the question of how J and M understood mathematics in a computational context related to chemistry: confidence and facility with qlathematics; intuitive knowledge; and learning.
Issue 1: Confidence and facility

J, the non-A-level student, suffers from a lack of confidence and facility in mathematics. But he reports that
doing the assignment significantly boosted his confidence: I: What did you think when you were given [the assignment]? J: For quite a long time not knowing how to go about it. I started early planning to get it out of the way and ended up spending more time than most people doing it. It was quite straightforward at first, but as it became more complicated I seemed to stray off the path, where other people were finding more direct routes, I'd sort of wander about and find different methods ................................It would have been more enjoyable with a bit more help on hand every so often, just to check, I spent a long time going down wrong paths, if I'd been steered more to the right direction I'd have had more fun doing it, done it much more quickly. I: Has having to try things out for yourself been a benefit, or a loss?
J: It encouraged me not to give up too easily, which is what I tend to do when I see maths ... I didn't do
maths, I can't do this-and don't really look at it, now I spend more time looking at it, think about what can I do. In contrast, student M, with double A-level maths and a strong programming background, describes doing the assignment in very different language: I: How did you decide to go about doing the task? M: In the same [question] sheet there was an example, showing how to get data in and how to play with it, so I followed that to start with, drew a graph ... first I had to remove the background
Page 3
count ["zero error"] which meant identifying which points were the background, finding the average of them and then subtracting that from all the data values ... then I had to turn the list of voltages, one-axis data ... the intervals between each reading were actually 467, so you had to thread together a second set of data, the values for the time.
Issue 2: Explicit and implicit knowledge, "intuition"

There is more commonality between the students on this issue: I: Did [doing the assignment] in any way improve your understanding of the underlying chemistry? Did you make use of chemical understanding?
J: A bit of both I think ... the knowledge of what was happening in the reaction would have suggested to
me, even if I hadn't seen the data, that it would have been exponential. Seeing the data and working with it helped to visualise better and understand what was going on in more detaiL .. I don't know which influenced which more. Similarly, for M the process was helped by his own experience with chemistry and mathematics: I: What was the process for fitting to the data? M: Partly thinking about, this is a biological experiment, what is there a faint chance of it actually doing? I: So you had in mind first of all the chemical processes? M: Urn, chemical things don't tend to do curves to the power five, anything like that. I: How did you know that? M: Just sort of an intuition, I'd never come across something like this before. I: But you had the sense that the reaction you were looking at had to have certain characteristics? M: Yes, either it was going to be linear, or some sort of quadratic, or something to do with exponentials. I: Where did that come from, that intuition? M: Mostly previous experience with experiments, since I've been doing chemistry really .......................... . The maths helps as well, you just generally get more of an idea.
Page 4
Issue 3: "What do you think you've learnt?"

I: Having done the assignment, what have been the gains for you, in terms of understanding the chemistry or the mathematics, if there were any? J: Certainly I've learnt to use the program much better, the different commands, the shortcuts, so the next project should be quicker. ... " I don't think I've learnt really any chemistry, I've learnt more about this particular reaction I suppose .......... The maths has improved in this particular area, the understanding of the shapes of the equations, which is one of the things that I lack, [it gave] lots of practice in seeing what the graph would look like without having to plot it, that's helped. 1's response here is typical of most of the interview group, in saying that he doesn't feel that he's learnt any mathematics or chemistry except "in this particular area". This seems to us a key point for investigating our triangle of relationships: that understanding. the relationship between knowledge domains is precisely about establishing connections in a sequence of "particular areas". M seems to express a sense of this, with his talk of "experience": I: What do you feel you've learnt using Mathematica, after all of this? M: You wouldn't be able to do this kind of analysis on a real data set without Mathematica, as an exercise you'd be restricted to a simple experiment with 10 or 20 data points, do things by hand.
I: In terms of mathematics, did you learn anything new? M: Not so much learnt, you get more experience, a better feel about how to make things fit, it'll be useful to use Mathematica now to do this kind of fitting, which I couldn't have done before.
Conclusions
At a superficial level it seems obvious that not having a background in formal mathematics, such as that given at A-level, made a difference to students' understanding of the assignment. Students without such a background may be forced to develop the capacities of perseverance, reflection, and self-guided instruction, focussed on a specific problem, that their more mathematically knowledgeable fellows may not be called on to use. In fact, through this harder work, our Chemistry colleagues have reported over the years that the non-A-level students often get ahead of the majority of their fellows, in developing good mathematical ways of working in the scientific context, because they are not able to coast along on the back of previous mathematical experience. These evaluation findings have important curriculum implications, in that these issues are of central concern to our colleagues in the Chemistry Department, given their aim of producing graduates who
Page 5
are mathematically-equipped to begin doing chemical research-which, in fact, is the first destination of about 50% of the graduates. The task that the students were given was principally about curve-fitting, and made relatively few demands on their chemical knowledge. However, as the Issue 3 extracts suggest, the nature and extent of the students' mathematical development is closely tied to the specificities of the context in which they meet the mathematics. This has important implications for mathematics education, since it raises issues in an "academic" setting that compare with our colleagues' work at the Institute of Education on mathematics in "vocational" situations (such as banking and nursing: Pozzi et aI, in press). We have now completed an extended "maths lab" observation of three pairs of students as they worked on another chemical-mathematical assignment (a simple Newtonian mechanical simulation of ionic interactions); plus, all these students were interviewed again. In analysing the full data set, we will again be looking for data about the triangle of relationships (mathematical knowledgechemical knowledgemathematics software), and the interaction of explicit and implicit knowledge, and the use of, and development of, "intuition". This analysis, we hope, will provide the basis for more substantial evaluation studies in subsequent years.
References
Pozzi, S., Noss, R. and Hoyles, C. (in press). "Tools in practice, mathematics in use". To appear in Educational Studies in Mathematics. Ramsden, P. and Templer, R. (in press). "A new approach to mathematical training for chemists". To appear in Education in Chemistry. Templer, R., Klug, D., Gould, I., Kent, P., Ramsden, P. and James, M. (in press). "Mathematics laboratories for science undergraduates". To appear in Rethinking the Mathematics Curriculum, edited by C. Hoyles, C. Morgan and G. Woodhouse. London: Falmer Press.
Page 6
A SURVEY OF PUBLIC IMAGES OF MATHEMATICS LIM Chap Sam & Paul ERNEST University of Exeter Abstract: This paper reports initial findings of a survey that aims to explore the range of public images of mathematics. Over 500 adults aged 16+ from all walks of life responded the short questionnaire given. Initial findings show that public images of mathematics and learning mathematics were given in the forms of propositions expressing opinions and views or in the form of metaphors. Five main categories of responses emerged from the analysis. They are (a) attitudes towards mathematics and its learning; (b) beliefs about respondents' own mathematical abilities; (c) descriptions of the process of learning mathematics; (d) epistemology and views of the nature (?f mathematics; and (e) values and goals in mathematics education. Some methodological issues and examples of each category are given and discussed in the paper. Introduction Mathematics is a mysterious subject, and a number of myths are associated with mathematics. These myths include commonly expressed views including: "mathematics is just computation", "mathematics is only for clever people (and males)"; "your father is a maths teacher so you must be good in mathematics too". Such myths and images are widespread, are seem to be present in many countries, and among all classes of people. Moreover, most of these myths are negative [Buxton, 1981; Ernest, 1996; Peterson, 1996]. It is a matter of concern that these negative images of mathematics might be one of the factors that has led to the decrease in student enrolment in mathematics and science at institutions of higher education, in the past decade or two. However, there are relatively few systematic studies conducted on the subject of myths and images of mathematics. We need an answer to the questions: What are the of the general public's images and opinions of mathematics? We need to ascertain how popular or unpopular mathematics is, before we can design measures to improve or promote better public images. Therefore, this study aims to explore the range of images of mathematics held by a sample of the general public. It also aims to investigate the factors which might influence or cause these images. However, this paper only reports the initial findings of the first part of this survey, which was carried out during summer 1997 in the UK, as part of a Ph.D. research study. The survey This study employs an interpretative design because images are personal constructs that involve both the affective domain [feelings and attitudes], and the cognitive domain [knowledge and beliefs], including metaphors and related images. The meaning of image is taken as the 'mental representation or mental experience of something that is not immediately present to the senses, often involving memory' (McLeod, 1987, p.497). Thus the term 'image of mathematics' refers to a mental picture,
Page 7
view or attitude towards mathematics, presumably developed as a result of social experiences, through school, parents, peers, mass media or other influences.
The sample
The sample consisted of 548 adults of the general public aged 16+ who came from all walks of life. To obtain a wide selection of the public, the location of data collection were public places such as a town high street, bus and train stations, an airport, and visitors to a cathedral, university campus and at a school open evening. We acknowledge that due to the constraints of time and resources, the sample is essentially an opportunity sample. (Although representatives for each empty or near empty sub-section of the sample stratified according to age band, gender, occupational grouping were actively sought, as far as possible: see Table 1). Only those who agreed to participate were sampled. The sample can be grouped into two main categories: First there is the public who are directly involved in mathematics education such as mathematics teachers and mathematics students (this was a small part of the sample, 76 in total). Second, there is the public who are not directly involved in mathematics education, including nonmathematics teachers and students. This second category of the public was subdivided into five occupational grouping based on one of the common social survey classification, the Social Class based on Occupation (also called the Registrar General's Social Class) (The Office of Population Census and Survey, 1990). The detailed distribution of the sample is as shown in Table 1. The sample was also classified (by self-report) into four age groups, namely, the youth group (age between 17-20); the young age group (age between 21-30); the middle age group (age between 3150); and the older age group (over 50 years old). The ages were grouped in such a way that the opinions of the sample might loosely reflect experiences of different stages in the evolution of the mathematics curriculum in England and Wales.
Page 8
Method
A short questionnaire using both open-ended and structured questions was designed to probe for the public's images of mathematics. It contains 10 questions which asked for respondents' liking or disliking of mathematics, their feelings when they thought of mathematics in school, their beliefs about mathematics learning, their views about mathematics education and their images of mathematicians. However, for this paper, only responses to two open-ended questions will be utilised. These asked the respondents to describe their images of mathematics and learning mathematics are discussed here. The responses were textual and were analysed qualitatively. The two open-ended questions are as given in Box 1 below:
Methodological issues/problems
Since the responses given were textual expressions, they were open to multiple interpretation. Some of the responses were given in the form of direct descriptions of attitudes such as 'mathematics is boring' or respondent's beliefs about their own mathematical abilities such as 'mathematics is misleading and confusing' [Respondent (R) 431]. However, many responses were also given in the form of a metaphor or simile such as 'mathematics is a nightmare' [RI8S] or 'maths is like Mt. Everest, difficult to climb but not impossible' [RS22]. This raises problems concerning the validity of our interpretation and data analysis. To overcome this problem, a few measures have been taken to validate our interpretation. First, all ambiguous data was eliminated unless independent confirmation of the interpretation could be made. For example, one response to the question on the image of mathematics was 'maths is a snail shell in the garden' [RI17]. It was discarded because it opens up to too many possible interpretations and it was not possible to get further confirmation because the respondent did not agree to be interviewed in the second stage of this study. On the other hand, another response given was 'maths is pen and paper' [R239] which is equally ambiguous as the first one. However, it was possible to reconfirm and clarify its meaning with the respondent because she agreed to take up the follow-up interview in the second stage of the study.
Page 9
Second, an attempt was made to classify the textual responses into as many categories as they can be represented in, because some responses contain composite views or mixed feelings. In order not to lose the richness of the data, each response was coded into one to three different categories or subcategories. For example, the response given: 'mathematics is a complicated but interesting subject' [R360] is coded into three subcategories: Code III : interesting; Code 411: a discipline or subject and Code 415: complexity. In the whole process of categorisation and re-categorisation, we realised that it would be naIve for us to think that we could avoid personal bias and personal values during interpretations of data. To minimize this personal bias and prejudice, the data was also cross validated by systematic triangulation. Firstly, the data was cross validated with four validators: two experts in mathematics education, one expert in quantitative research and one postgraduate research student. Their ages range from 29 years old to 60 years old. All of them have lived in Britain for more than 10 years, ensuring that they are familiar with the British culture, society and language. They were given the list of categories with verbatim examples taken from the data. We discussed and readjusted some of the categories according to our interpretations of the data. We then come to a compromise that resulted in a modified list of categorisation of the data. This process was repeated with each of the four validators. Secondly, parts of the data together with the categories were validated by participants in a national mathematics education conference (BSRLM). The participants were given the list of categories and 20 sample responses in the form of metaphors or descriptions of the images of mathematics. After an explanation of the category chart (Figure 1, earlier version), the participants were divided into small discussion groups and given 15 minutes to categorise the 20 sample data according to their own interpretations. They were encouraged to create new categories if they felt the suggested list of categories was not adequate for the data given. After the group discussion, the category assigned by the researcher and the participants were compared. Only four items out of the 20 sample data were not matched. A few new categories such as 'impossible' emerged from the discussion and many more multiple categorisations were suggested for each item of data in the sample. Therefore the category chart was further modified and readjusted in the light of these suggestions. As a result of these two layers of validations, the final chart of categories for the images of mathematics of the sample was developed and is shown in Figure 1.
1 0
Page 10
Findings and Discussion Initial findings show that public images of mathematics and learning mathematics were given in the forms of propositions expressing opinions and views or in the form of metaphors and similes. Five main categories of responses emerged from the analysis. They are (a) attitudes towards mathematics and its learning; (b) beliefs about respondents' own mathematical abilities; (c) descriptions of the process of learning mathematics; (d) epistemology and views of the nature of mathematics; and (e) values and goals in mathematics education.
Category 1: Attitudes towards mathematics and its learning
Initial analysis shows that many respondents expressed their images of mathematics in the form of descriptions of attitudes, feelings or emotions that they had when they thought of mathematics or what the questions reminded them of Descriptive statements such as 'mathematics is difficult' [frequency(t)=68]; 'mathematics is boring' [f=59] or 'mathematics is interesting/rewarding' [f=62] are the three most common expressions. Metaphors which show positive images such as 'mathematics is like playing with my children, never tiresome' [R526] or negative images such as 'like eating nails -hard and painful' [RIll] are also commonly expressed. As shown in Figure 1, over 44% of the entries indicate some kind of attitude, feeling or emotion. They range from positive attitudes such as 'mathematics is fun and exciting' [RI70] to negative attitudes such as 'mathematics is dull, boring complex' [RI82]. Others emphasise the importance of mathematics such as 'mathematics is important for everything' [RI75] while yet others see mathematics as 'irrelevant' [R053] and 'a lot of things which I will never use' [R059].
Category 2: Beliefs about own mathematical ability
Five percent of the responses reflected beliefs about the respondents' own mathematical abilities and experiences. Some believed that mathematics is difficult but possible to achieve success in, examples mathematics is 'like Mt. Everest, difficult to climb but not impossible' [R522]. Others hold the opposite view, example, 'mathematics is difficult and [I] find hard to cope with' [RO 11]. A number also found mathematics 'incomprehensible' [R027] or 'misleading and confusing' [R43I].
Category 3: Descriptions of the process of learning mathematics
Almost 14% described their images of mathematics in terms of the process of learning mathematics. Example, maths is, 'a skill you need to learn' [R209]; 'problem solving, explaining physical processes'[Rl13]; or a 'voyage of discoveries' [RlI6] and as exploration, for example, maths is 'like the arctic - unattractive but adventurous'[Rl08]. Many of the sample indicated that maths is a hierarchical process like 'a seven course meal, one theory leads to another' [R527] or involved logical thinking such as 'logical stimulation' [RIOO] or mental work such as maths is 'a subject to
Page 111\
test the mind' [R468]. Mathematics is also viewed as process of 'hard work' [R034], or effortful endeavour as 'something required concentration-satisfy when right' [R090]; a repetitive process, example, maths is 'repetitive, structures and logical' [R105]. Among these eleven subcategories, logical thinking, mental work and problem solving were the three most common suggesting that for those who relate mathematics to learning, it is taken as a cognitive process of logical and analytical thought to solve problems.
Category 4: Epistemology or the nature o.fmathematics
Almost 30% of the responses corresponded to this category, making it the second most prominent category. For example, mathematics is identified with 'numbers and equations' [R005]; rules and procedures, pattern and structures. Mathematics is also viewed as a practical tool, a model, a language, a science or a discipline of study. The responses corresponding to this category are made up of a wide variety of constructs that related to epistemology, content and nature of mathematics.
Category 5: Values and goals in mathematics education
Over 8% of the responses referred to the goals and values of mathematics education. Eight subcategories emerged from the data. The most common category concerned the element of mystery in mathematics, example maths is 'like a woman - full of intriguing mysteries' [R167] or 'like swimming in the dark' [R442]. It is interesting to note that all responses in this subcategory were given as metaphors. The second most common category viewed mathematics as a challenging activity, example, maths is 'challenging' [R487] or 'a challenge' [R237], 'fun when everything works out but remains a challenge' [R470], or a 'challenging subject -- intellectually satisfying' [R120]. There were also responses that show the appreciation of the values in mathematics or the beauty of mathematics, example, 'mathematics is clean and reliable' [R061] or 'mathematics is like a sunset- unique and beautiful' [R168]. Conclusion This initial analysis shows that the public's images of mathematics, at least in this sample, are closely related to their attitudes and feelings towards mathematics. The majority of them found mathematics difficult, boring or rewarding, and this was their most significant perception. Although many related their images of mathematics to the nature of mathematics, that is, to epistemology or content, by identifying maths with numbers and equations or rules and procedures, about 14% of them related maths to its process of learning. Less than 10% of responses related to values in mathematics education or reflected their beliefs about their own mathematics abilities. Further analysis in terms of
I L
Page 12
gender, age groups and occupational grouping as well as the possible factors that might have influenced these people's images of mathematics are underway.
References:
Buxton, L. (1981). Do You Panic About Maths? London: Heinemann Ernest, P. (1996). Popularization: myths, massmedia and modernism. In A. 1. Bishop, Ed., The International Handbook of Mathematics Education, Dordrecht: Kluwer Academic, Vol. 2, pp.785-817. McLeod, W.T. (Ed.)(1987). The new Collins dictionary and thesaurus in one volume. UK: HarperCollins. Office of Population and Census and Surveys, and Employment Department Group (1990). Standard Occupation Classification, Volume 1: Structure of the classification. London: HMSO. Peterson, I. (1996). Search for new mathematics. The Mathematics Forum, webmaster@forum.swarthmore.edu. Error! Hyperlink reference not valid. ivars.html>
Page 13\3
Page 14
MODELLING, STRATEGIC QUESTIONING AND THE LAWS OF STUDENT REASONING IN A-LEVEL MECHANICS
Stuart Row lands, Ted Graham and John Berry The Centre for Teaching Mathematics University of Plymouth
The Newtonian concept of force is a functional quantity that explains changes in motion - force as a relation between two bodies. Many A -level students, on the other hand, conceive force ontologicaliy as a property that the body possesses. There is much research that suggests that these ontological beliefs are not acquired through experience of the physical world prior to learning mechanics, but may be formed when the student is asked to consider qualitative examples of force and motion for the first time. This paper reports on two possible laws of student reasoning in mechanics based on the results of a pilot-study.
INTRODUCTION
This paper is a continuation of the proceedings paper (Rowlands et ai, 1997) on the Socratic method as a teaching strategy that challenges 'misconceptions' and facilitates the construction of the Newtonian system within the students zone of proximal development, and reports on a study that suggests two laws of student reasoning in mechanics. 'Typifying the view of Vygotsky', Davydov (1988) argues that the method of (developmental) teaching should be based on its content - namely the logical structure of the subject being taught. Taking our cue from Vygotsky and Davydov, we begin with the logical structure of Newtonian mechanics.
LOOKING AT BOTH ENDS OF THE ZPD: THE STRUCTURE OF NEWTONIAN MECHANICS AT ONE END, AND THE NATURE OF INTUITIVE IDEAS AT THE OTHER
We do not experience force! What we do experience, however, are macroscopic objects interacting with macroscopic objects, including ourselves as macroscopic objects. When we see two objects colliding, we see the changes in motion as a result of the collision. What we don't see, however, is the force of interaction acting between the two objects. When we see a ball thrown into the air, what we see is its trajectory as a result of the gravitational pull of the earth (another object). What we don't see, or even experience, is the gravitational pull of the earth. If you are in freefall enclosed in a capsule of some kind, then you will experience the sensation of weightlessness, even though you are not weightless:
Page 15
1 5
When you travel around a bend, you do not experience the centripetal force acting on you towards the centre of rotation. What you do experience, however, is the sensation of being pulled outwards (an experience of the imaginary 'centrifugal force'). If you were hit on the head with a hammer, and you momentarily experienced an arrow labelled 100 Ns, then that experience may be attributed as a form of concussion! The point is that force in mechanics is not something we can experience. We experience the actions of other bodies, but force in mechanics is a concept, a functional quantity, that explains changes in motion due to the actions of other bodies. Force is not a property of an object, it is not a quantity that the object possesses. Rather, it is a relation between two bodies (either in contact, such as friction or normal reaction, or at a distance, such as gravity) and that relation is exactly specified by the laws of motion. The laws of motion are not empirical laws that can be discovered by measuring the variables in an experiment (a very important pedagogical point for those keen on an experimental or 'hands-on' approach. See Rowlands et aI, 1999). They are not empirical generalisations such as 'all ravens are black', but express transfactual tendencies (Chalmers, 1982). In explaining the way in which physics speaks of the world, Chalmers (1982) states:
Let us take as an example Newton's first law of motion, the law that Alexander Koyre described as the explanation of the real by the impossible. Certainly no body has ever moved in a way that perfectly exemplifies that law. Nevertheless, if the law is correct, all bodies obey it, although they rarely get a chance to show it. The purpose of experimentation is to give them a chance to show it. If Newton's laws are 'true' they are always 'true'. They are not true only under experimentally controlled conditions. If that were so we would not be justified in applying them outside of experimental conditions. If Newton's laws are true they are always true, but are usually accompanied by the simultaneous action of other tendencies. If Newton's laws correspond to
Page 16
anything it is to transfactual tendencies, which are very different from localised states of affairs such as cats being on mats (p. 155). The laws of motion are axioms and not empirical generalisations. They do not by themselves tell us anything factual about the objects of the world, but they do speak indirectly about the objects of the world by stating the precise way in which it is possible to describe the world by these means (Wittgenstein, 1961). The coherence of Newtonian mechanics consists in the way it accounts for all contingencies. The laws of motion define force in a way that stipulates how a phenomenon shall be treated by the system, and the system is unified in the sense that it can accommodate all accountable phenomena within the limitation of its domain (macroscopic objects with speeds that are not comparable with the speed of light). To quote Wittgenstein (1961): Newtonian mechanics, for example, imposes a unified form on the description of the world ........Mechanics determines one form of description of the world by saying that all propositions used in the description of the world must be obtained in a given way from a given set of propositions - the axioms of mechanics. It thus supplies the bricks for building the edifice of science, and it says, 'any building that you want to erect, whatever it may be, must somehow be constructed with these bricks, and with these alone'. (Just as with the number-system we must be able to write down any number we wish, so with the system of mechanics we must be able to write down any propositions of physics that we wish) (proposition 6.341). Mechanics is an attempt to construct according to a single plan all the true propositions that we needfor the description of the world (proposition 6.343, author's emphasis). The way mechanics speaks of the world is unified, and the coherence of mechanics consists precisely in the way the system is a unified form of description. Student intuitive reasoning, on the other hand, lacks the coherence of the Newtonian system - it is not unified but fragmented (Rowlands et aI, in print). Newtonian mechanics is consistent in its account of different phenomenon, whereas student intuitive reasoning tends to lack consistency - it differs with respect to each phenomenon. Students may regard force and motion according to how they conceive the dominant features of the phenomena under investigation: the force acting on a car braking is in the opposite direction to the motion, yet the force acting on a ball thrown upwards has to be upward because the ball is going upwards. We do not experience force, but we do experience the motion of bodies, and our ideas of force are sometimes muddled with what we consider to be the dominant features of motion. To understand mechanics in the way it describes the world, to explain phenomena qualitatively according to the Newtonian system, requires thinking in the abstract! With facilitation by the teacher, the class has to construct for itself the Newtonian system as a unified form of
Page 7 17
description - a sharp contrast to the fragmentary responses given by students as reported in much of the literature. To explain the world in accordance with mechanics as a unified form of description is to 'explain the real by the impossible' - that if objects were given a shove on a frictionless surface then they would move in unified motion, that the laws of motion would be invariant in a possible world of no gravity, etc. ldealised abstraction are the rules upon which we have to think in terms of mental models, and these rules are the laws of motion (Rowlands et al, 1999). Students may appear to have little difficulty in dealing with idealised examples that demand a quantitative response (e.g. Calculate the
acceleration of a particle down a frictionless plane inclined at 30 to the horizontal). The real difficulties
appear when students are asked to explain the world qualitatively within the constraints of idealised abstraction (e.g. Two particles, one with a mass several times greater than the other, are released
simultaneously from the top of an inclined plane. Which particle will reach the bottom first?). Of course, the
teacher can always set up the conditions that would reveal the answer. However, that would still leave the class to understand why both particles would reach the bottom at the same time. The asking of concept and parallel questions presupposes a meaning that is only relevant to idealised abstraction. Each question is framed in a way that reveals the consistency of the Newtonian framework, but the raising of each question is determined by the responses to the previous question. The appropriation of meaning is structured within idealised abstraction, but the class is continually invited to construct that meaning as the students attempt to instantiate intuitive ideas to make sense of the phenomena as presented by the teacher. The teacher should not attempt to refer to these intuitive ideas as part of a 'scaffold' to facilitate a Newtonian understanding, but should raise anomalies that have the same explanation under the Newtonian system. In this way, Newtonian mechanics may be seen as a semiotic system - much more meaning is conveyed than is actually stated in the asking of concept questions, and cognitive strain will result as the student attempts to infer the meaning of the questions. 'Misconceptions' or 'alternative frameworks' may be considered as initial attempts to construct a meaning, and may be resilient to change given the cognitive strain in forming them. Concept questions may promote cognitive conflict (Graham & Berry, 1994), in the sense of the gestalt-shift from 'seeing that' (in terms of the intuitive ideas instantiated) to 'seeing as' (in terms of the Newtonian model).
Page 18
CONCEPT QUESTIONS AS A DIAGNOSTIC TOOL AS WELL AS A TOOL OF REMEDIA TION - THE LAWS OF STUDENT REASONING IN A-LEVEL MECHANICS Force, as defined
within the Newtonian system, is a relation between two bodies. Many students, however, conceive force ontologically - as a property that the body possesses (Nersessian, 1992). These ontological beliefs are not acquired through experience of the physical world prior to the learning of mechanics (force cannot be experienced), but may b~ formed when the class is asked to consider qualitative examples of force and motion for the first time. Conceptual changes of force and motion requires the construction of new concepts, together with the challenge to emerging intuitive beliefs, through strategic questioning. However, by challenging emerging intuitive beliefs we can reveal the laws of student reasoning in mechanics. Similar to the 'experimental-developmental' method of Vygotsky which calls for an experimenter to intervene in some developmental process in order to observe how such intervention changes it (Wertsch, 1985), concept and parallel questions can be used to evoke responses so that we can analyse psychological processes that occur in mechanics instruction. In other words, within the zone of proximal development of a mechanics class, if we are to understand the cognitive abilities of the class as a process then we have to instigate that process by interaction with the class. The pilot-study with the HITECC (engineering foundation year) mechanics class, and informal intervals with sixth-formers, has suggested two laws of reasoning: Some students can develop their personal intuitive beliefs of force and motion with remarkable consistency and coherence (regression from the target concept), until an anomaly contradicts their ontological status of force . The greater the regression, the greater the cognitive conflict necessary.
Concept questions not only reveal the intuitive beliefs of the students, they can also arouse the students minds to life, sometimes with a remarkable coherence in their attempts to defend their beliefs. The three students from the Hitecc class pilot-study were able to respond consistently to the questions that were parallel to the concept question of the force acting on a vertically thrown ball: Teacher. But when I throw a ball and place my hand in my pocket... .. Student 1. Yeah, but the force has comefrom you but is decreasing. Teacher. It's decreasing? Student 1. Yeah, as its losing velocity, its going to slow down and the force of gravity is going to be greater which is forcing it downwards. Teacher. But if I am on a hull of a spaceship and I throw the ball, how is it going to go? Student 1. It is going to go straight. Teacher. But what is its motion like when it is going straight?
I"J Page 19
Student 1. Uniform. Teacher. Is there aforce pushing it? Student 1. Not after it has left you There appears to be a sigh of relief from the class. Teacher. [looking at students 2 and 3, who seemed to have shared the same ideas as student 1] So where is this force coming from then? Student 2. Its own mass. Student 3. l'd say it is from you, but decreasing. Class appears to be a little unsettled at this point. [Student 2 appears to have a medieval impetus viewpoint, while student 3 appears to have an Aristotelian one. It was tempting to abandon the discussion since the whole class with the exception of the two students appeared to have grasped the point. However, this may have been a consensus within the class to 'keep me happy' so as to move on. Moving on was a temptation because subsequent concept-questions could always be related to the previous phenomena considered. Nevertheless, I had attempted to 'crack this nut' as it might have revealed a successful strategy]. Teacher. But why is the force from me, from the hull of the spaceship ............ . Student 3. There is no resistance ... Student 2. Oh, I know, its the force stored in the mass. Teacher. Force stored in the mass? In fact that is quite often given, even by Aristotle [a mistake, I should have said the medieval impetus school]. Aristotle argued that force is a property of the object. Student 2. The force given by the thrower goes into the ball - into the mass - and in that situation, stored in the mass, and because nothing is stopping it - it is in a vacuum - it just carries on. But if another force is applied to it, it will change direction [hand gesture of a projectile's trajectory]. Student 4. [who has been waiting a long time to respond]. The only force acting on a thrown ball is gravity. If a ball is in uniform motion then no force is acting on it. Student 5. The force is gone.
Students 2 and 3 had set themselves apart from the rest of the class, the majority of which appeared to have little difficulty in giving a Newtonian response to the original concept question. Subsequently, the three students became very consistent (and entrenched) in their arguments. For example: If there is no such force [pushing a projectile] then why does it go to the top of the arc before gravity takes over? If a ball hits a wall, then the dent in the wall is caused by the force in the ball. There is a force opposing the motion of the car braking, namely friction; but the force pushing the car gradually weakens - and that is why the car slows down until it stops. When the car stops, the frictional force and the push are in equilibrium. I raised the example of the ball held between two fingers on a train travelling in uniform motion and asked what was the speed of the ball relative to a stationary observer and if there were any horizontal forces acting on the ball. There were facial expressions of cognitive conflict from
Page 20
students 2 and 3. I had raised the question almost by accident, as I had just remembered the example raised by Gallileo of the sailor in the crows nest who drops a ball - does the ball land in front of the mast, at the foot of the mast or behind the mast? Throughout the rest of the course the two students were able to provide a Newtonian account of complicated qualitative models. No one has observed or experienced uniform motion in the total absence of force; nonetheless, the anomaly of the ball in the train is the closest that one can consider uniform motion in the absence of force. I was able to formalise the first law of motion and to move the class collectively from the idea of uniform motion in the absence of force to uniform motion in the absence of a net (resultant) force: Horizontal uniform mition requires no force Horizontal uniform motion and horizontal forces ~ net horizontal force = 0 In the above discussions, the students were becoming more and more consistent in their developing argument that force is a property of the object and that this force is given to the object by an agent (a thrower). Their developing argument had assimilated each parallel question according to the original notion that a thrown ball moves up because the force required to overcome gravity is the force that was given by the thrower. However, there appeared to be expressions of cognitive conflict on the faces of two students when the anomaly of the ball on the train was asked. Subsequent informal one-to-one interviews with a dozen preinstructed sixth-formers suggests that if this question is asked after a series of related parallel questions, then cognitive conflict will occur (perplexed facial expressions, minutes of silence, fidgeting). On one occasion a student refused to see the relevance of the question after giving the correct answer. Once cognitive conflict occurred it appeared relatively easy to prompt the student to explain why a vertically thrown ball slows down. However, on three occasions I had given the ball on train question first. In each case the question was answered correctly (with no apparent cognitive conflict), but subsequent parallel questions were answered incorrectly. This suggests that the ball on train question has an 'impact' only if the question is
asked after a series of parallel questions. It seems as though the question cannot sustain a developing
'force is the property of the object' argument. In other words, the answer to the question undermines the belief that the ball has to have a force in order to maintain uniform motion. We are now involved in a research project that includes clinical interviews that would hopefully verify the two laws of reasoning in mechanics: The more parallel questions that can be 'assimilated', the greater the entrenchment and hence the greater the need to induce cognitive conflict. Cognitive conflict can occur if a question undermines the students ontological status of
Page 21 1
force. If these two laws are verified, then the next stage of the project will be the consideration of the laws in a classroom situation.
REFERENCES
Davydov, V. V.: 1988. 'Problems of Developmental Teaching: The experience of theoretical and experimental psychological research', Soviet Education, 30(9), Sept. Chalmers, A.: 1982. What is This Thing Called Science? Second Edition. The Open University Press, Milton Keynes. Graham, E., and Berry, J.: 1994. 'Concept Questions in Mechanics,' Staff and Educational Development Association paper no. 87 (Dec.): Innovations in Mathematics Teaching, ed. K. Houston. Nersessian, N.: 1992. 'Constructing and Instructing: The Role of Abstraction Techniques in Creating and Learning Physics.' Philosophy of Science, Cognitive Psychology, and Educational Theory and Practice, ed. A. Duschl and R. Hamilton. State University of New York Press, Albany. Rowlands, S., Graham, E.
& Berry, J.: 1997. 'The Socratic Method of Strategic Questioning to Facilitate the Construction of the
Target-Concept Within the Students Zone of Proximal Development', BSRLM Bristol conference proceedings, 15 Nov. Rowlands, S., Graham, E. & Berry, J.: 1999. 'Can we Speak of Alternative Frameworks and Conceptual Change in Mechanics.' Science & Education, 8(3), May. Wertsch, J.: 1985. Vygotsky and the Social Formation of Mind. Harvard University Press, London. Wittgenstein, L.: 1961. Tractatus Logico-Philosophicus (First English edition, 1922). Routledge and Kegan Paul, London.
Page 22
INVESTIGATORY APPROACHES IN MATHEMATICS TEACHING: IMPLICATIONS FOR TEACHER TRAINING IN INDIA Dr. Viiayalakshmi Chilakamarri. King's College, London, Visiting Academic from INDIA.
In the present age of Science & Technology it is very much essential to develop the process skills, experimental skills, thinking skills and the skills of exploration among the school children. Mathematics is an appropriate subject to inculcate and develop these skills through Investigatory approach. So it is necessary to know to what extent Investigatory Approach is adapted while advocating it in teaching mathematics, in the Initial Teacher Training Institutes and its impact on teaching mathematics in the Secondary Schools. An attempt is made to study in detail how the Investigatory Approach is adapted in the Initial Teacher Training Institutions at the University level and how is it practised at the Secondary school level to improve the quality of mathematics education. My observations,findings and their implications to the Teacher Training in India are put before you for discussion. Any suggestions offered are highly appreciated.
1. Introduction
Recent Research findings in different parts of the world are stressing the need of adoption of strategies that provide opportunities for children to think rather than use the product of thoughts, to do mathematics instead of reading and listening mathematics. In India there is a great dissatisfaction among mathematics teachers and teacher educators about the way mathematics is being taught particularly at primary and secondary school level. We feel that we are not providing sufficient opportunities and environment for the pupils to think mathematically. In this context I came to England on 29 July 1997 to study the Instructional Strategies and Information Technology adopted and advocated in Initial Mathematics Teacher Education Courses in Teacher Education Institutions in UK and their impact on the quality of teaching and learning mathematics at Secondary school level and their relevance to the local conditions in India. Of all the strategies that were used in the PGCE, the investigatory approach caught my attention since it is the approach that is based on Ausubel & Robinson's (1971) psychological aspects of mathematics learning namely 'Meaningful Discovery Learning.' When I came here there were so many questions on my mind and I was very keen to find the answers for them during my observation of the sessions. They are 1. What are the features of an investigatory lesson (including the role of the teacher)? 2. What factors facilitate trainees' use of investigatory approaches? (a) presence of approach in placement school (b) beliefs in philosophy in placement school (c) beliefs in trainees (d) university sessions embodying approach in incorporating discussion of lesson approach
BSRLM 28/02/98
Vijaya-Investigations
Page 23
(e) university sessions embodying investigatory approach These questions worried me because the way that we follow this approach in India is not systematic or well structured. At the Primary School level we call it 'Activity Based Learning'(ABL), not an investigatory approach.
2. Current use of Investigatory approaches in Indian Primary & Secondary Schools In ABL, the
teacher provides an open-ended activity to all the students in the class by setting up a task based on the set mathematics syllabus. At the end of the task the findings of the students are summarised. To mention few 1. Post-Office activity: It is a whole class activity. To teach about Indian currency and moneytransactions' children are allowed to take as much toy-currency as they want (initially a particular amount). With the amount picked up they plan to buy the stamps and envelopes. In a simulated situation they do the money transactions. Those who complete the buying they plan the other combinations possible with the same amount and make a list. 2. Kiddy Bank Operations:- It is based on actual money transactions. This activity starts at the primary school and continues till they leave school. High school students manage the bank operations (issuing the pass books, withdrawal forms, maintaining the cash book ... ) and all the primary and secondary school students use the bank services. 3. School Stationery Shop:- It is run by the students. Goods are sold at concession rates and the Profits are used for the benefit of the students. 4. Activities using tamarind seeds: Students explore innumerable mathematical ideas using this activity. This is used at primary school level. a) Splitting the picked up seeds in different numbers of groups in different ways. b) Splitting up the picked up seeds to show the subtrahend, minuend and the remainder in different ways. c) Showing the picked up seeds as a product of two or more factors in as many ways as possible. d) Distributing the seeds picked up among different students of their choice equally and asking them to explain how many seeds were distributed, how many each got and how many are left. e) Ratios, fractions, percentages, patterns, arrangements and so on. 5. Planning school bus routes to teach route matrices (at Secondary school level). 6. Wiring a building - for teaching incidental matrices (at Secondary school level). 7. Activities leading to exploration of relations and generalisations - before teaching theoretical proofs in geometry (at Secondary School level). It is somewhat like guided discovery.
3. Current use of Investigatory approaches in Teacher training in India

The Investigatory approach is not directly adopted in teacher training. All institutions advocate it but few give direct experience of it to teacher trainees. It is insisted that during micro-teaching, every student
BSRLM 28/ 02/98
Vijaya-Invesligalions
Page 24
teacher has to teach at least one lesson using this 'activity based learning'. Feed back is given to each student teacher for improvement. In this context I visited Schools of Education (mathematics) in three universities in England namely King's College London, Cambridge University and University of Nottingham to observe how student teachers taught the Investigatory approach so that they would be able to use this approach in teaching mathematics to the school children. In King's College I attended all the PGCE sessions in the School of Mathematics Education. I prepared an Observation Schedule and recorded my observations.
4. My observation of Investigatory approaches during secondary teacher training in three British Universities 4.1
Examples of sessions observed:The following are the examples of the observed investigatory approaches in three different universities in England. 1) On 21 Oct 1997 from 9.30 a.m. - 12.30, a session with title 'Using and applying mathematics & investigations' in PGCE of Gill Close was observed in King's College, London. Gill started the session with seven aims regarding the abilities that the trainee is expected to acquire after the session along with three general aims and links such as linking with their own lesson preparation, identifying students' errors and linking to instrumental and relational learning. The task was 'Looking at Number Patterns using Octagon Loops' that led the trainees to investigate and find a rule to calculate the number of free edges in the next loop before making the loop. Five types of materials like octagon tiles, octagon template, starter sheet ... are used and in the process student teachers participated in seven activities. In this approach teacher exposition, whole class discussion, individual practical work & informal discussion, investigative work, discussion in pairs, small group discussion, student presentation and ultimately whole class discussion techniques were employed. Though it was a session on 'Investigations', an investigatory approach was followed. At the end of the session to make the trainees realise the importance of an Investigatory approach in using and applying mathematics and to know to what extent the trainees would be able to use it in their teaching, questions of the following type were asked. Why is it important to work investigatively? What should we teach through investigating? Think how you can include an investigative approach into your pair teaching in school within the next 3 sessions. 2) On 24 Nov. 1997 from 9.00 a.m.-12.30, a session with title 'Getting to know Calculators'Investigation on Calculators, in PGCE of Kenneth Ruthven was observed in the School of Education, University of Cambridge. Calculators with different calculating logic and work sheets were given to all the PGCE students and certain tasks were given for exploration of the use and applications of calculators. Cognitive conflict was used in the investigatory approach. The session started with a common starting
BSRLM 28/02/98
2
From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author Page 25
point 'Getting to know calculators' with teacher exposition. Then the investigational activity was given. The task given was: 'predict the result 3 + 4 x 5 = when you use your calculator and test your prediction, predict the result 80 + 15% = on your calculator and test your prediction, What are the largest and the smallest numbers can you make on your calculator, What happens when you subtract the smallest from the largest number?, multiply the smallest by the largest number on your calculator, explore these features and other features of your calculator,' and so on. Since the calculators use different logic, individuals got different answers. Students thought about individually and wrote their responses. It was followed by pair discussion with written responses, followed by small group discussion and subsequently whole class discussion. Observations were discussed and generalisations were made. 3) On 2 Dec. 1997, from 9.00 a.m. - 12.30 and from 1.30 p.m. - 5.00 p.m., sessions on 'Exploring Statistical Data using Computers' in PGCE of Malcolm Swan were observed in the School of Education, Nottingham University. Sessions were started with teacher exposition. Statistical data were given on different Real programs to all the trainees to explore, e.g. "Accidents", "Nations", "Sports", "Shirts". Work was differentiated due to trainees' expertise. Additional resources (suggestions for a 'lead lesson', work sheet for students who get stuck ... ) were provided. Trainees worked on computers in pairs. They explored the statistical data in various ways, including using statistical graphs, and made many interpretations and suggested solutions to the problems logically with statistical evidence. Techniques like discussion in pairs, whole class discussion, students' presentations were used. 4.2 Summary of University Observations:After observing six sessions (three in King's College London, one in Cambridge and two in Nottingham Universities), I found tentative answers to my fIrst question, 'What are the features of an investigatory lesson?' Some important features that I observed were: a) Cognitive Conflict was promoted by providing situations leading to different reasoning. b) Always there was an attempt to know how the trainees were thinking, what was happening in their minds, why they selected a particular strategy, why some of the trainees used alternative strategies. c) There was a continuous diagnostic assessment of trainees' understanding and misconceptions, which informed interventions. d) The focus was on processes and application of the content, but not the content alone. e) An attempt was made to know how the trainees were relating their previous mathematics knowledge to unfamiliar and challenging situations. f) Through individual, pair, group and class discussions the teacher observed how the trainees explored the given data and the mathematical reasons behind them. g) Throughout the sessions trainees were facilitated in thinking mathematically. h) In some sessions, in addition to following an investigatory approach, such an approach was also advocated and at every stage questions were asked to make the trainees realise why a particular
BSRLM 28/02/98
Page 26
Strategy was used. i) In all the sessions I observed the teachers behaved as an investigator, enabler, facilitator, questioner, listener, observer, a positive evaluator and a communicator.
5. My school visits
In order to obtain an answer to my second question 'What factors facilitate trainees' use of investigatory approaches?', I visited three schools in which King's College trainees were placed. These schools were categorised as A, B and C according to the schemes that they are following. In school A, CAME, GAIM and DIME are being followed. In school B, SMILE is followed, whereas in school C, SMP, STP, Task mathematics and Causeway books are followed. I gave a questionnaire to evaluate the impact of College and phase A on Phase B (Phase A is 1st placement- pair teaching or group teaching and Phase B is long teaching practice for the whole class) to the PGCE (mathematics) students of King's College London. a) In school A I saw: 4 SMILE lessons, taught by 3 different teachers, each with 2 King's college trainees supporting them, and a meeting of the mathematics mentor with the four trainees. b) In school B. I saw: 5 whole class lessons taught by 3 different teachers, each with 2 King's College trainees supporting them and a meeting of the mathematics mentor with the 4 King's College trainees. c) In school C. I saw: 4 whole class lessons taught by 2 different teachers each with 3 King's College trainees supporting them; One lesson in which the 3 King's College trainees were working with 3 pupils, 1 teaching, 1 helping and 1 noting the observations; and a meeting of the mathematics mentor with the 3 King's College trainees part of which evaluated the trainees' teaching of the 3 pupils. 1) The 8 whole-class lessons were focused on content, using either text books or work sheets. The ninth lesson was on revision of the previous lessons. Only features b), c) and part of i) were observed as teacher trainees associated to evaluate and assist pupils. 2) In the SMILE lessons some pupils were doing investigations but working alone, infrequently benefiting from discussion with pears or teacher intervention. During teacher / pupil interaction features b), c) and part of i) were again observed. 3) In the mentor sessions, one focused on planning materials for SMILE, another on teaching materials including using games, and the other on diagnosing weaknesses of the 3 pupils taught who were very able. 4) The commitment of the schools to CAME, Task Mathematics and SMILE suggests intervention investigatory approaches as do some comments and actions of mentors. 6.
Findings from Questionnaires

The information obtained from the 34 filled in Questionnaires by King's College PGCE (mathematics) reveals that
BSRLM 28/02/98
1
a) 21 trainees listed 3 important misconceptions and 13 listed only one or two that they identified either during their teaching or from oral or written diagnostic assessment, 7 out of 34 did not suggest measures to be taken to over come. b) Nineteen of them have given diagnostic assessment for the lessons ranging from 1 - 4, others have listed the misconceptions from their observations. c) 20 of the trainees mentioned either they used cognitive conflict or they are planning to use it. , Probability', 'Estimations' are few examples where they used cognitive conflict Some refereed CAME lessons where they used cognitive conflict. d) Four trainees stated that they cannot say the best lesson among what they taught where as 11 trainees attributed the investigatory approach (directly or indirectly) for their best lesson and rest of them attributed to the teacher exposition and explanation. 'Symmetry', 'Decimals and numbers on a scale', 'Enlargements' are few examples where trainees followed investigatory approach.
7. My dilemma
Whether it is India or England, what I have observed is the gap between the expectations in teacher training institutions and happenings in the schools. Still the following questions are worrying me. a) It is universal and everybody accepts the merits of the Investigatory approach. Then why is it not given enough emphasis in school teaching? b) Is it because of the presence of the long time existing approaches in the schools? c) Is it because of the belief in philosophy that is crept in the school? d)Is it because of the belief of trainees in the approaches through which they were taught at the school? e) Is it because of the University sessions are very idealistic? I have my own opinions about them but I do not have empirical evidence. So I would like to share your ideas, experiences and opinions in this regard. I would welcome feedback and suggestions for this research.
8. Implications for teacher training in India

On the basis of this discussion I would like to set assignments in India on the following issues. 1. Planning some investigatory approaches initially for Secondary school Teacher training to teach topics like arithmetic progressions, permutations & combinations and data handling systematically evaluating their take up by trainees in schools. 2. Planning some investigations for Primary school Teacher training to teach topics like exploring rational numbers, irrational numbers and decimal fractions. 3. Planning investigations based on patterns, mathematical structures like 'jumping frogs' for recreation and to create interest in algebra. Instead of Computer package in India we may use red and green counters.
***** *
BSRLM 28/02/98 Vijaya-Invcstigations
Page 28
Geometry Working Group

A report on the meeting at the King's College, University of London, 28th February 1998 Convenor: Keith Jones, University of Southampton, UK
Theoretical Frameworks for the Learning of Geometrical Reasoning With the growth in interest in geometrical ideas it is important to be clear about the nature qf geometrical reasoning and how it develops. This paper provides an overview of three theoretical frameworks for the learning of geometrical reasoning: the van Hiele model of thinking in geometry, Fischbein's theory offigural concepts, and Duval's cognitive model of geometrical reasoning. Each of these frameworks provides theoretical resources to support research into the development of geometrical reasoning in students and related aspects qf visualisation and construction. This overview concludes that much research about the deep process q{the development and the learning of vi sua lis at ion and reasoning is still needed. It seems that while for most of the twentieth century the mathematical literature has been predominantly algebraic, a growing interest in geometrical ideas has been stimulated by the development of powerful computer-based geometry and visualisation packages. The prediction is that such computer technology will have a significant positive influence on the progress of mathematics (National Research Council 1990, Science and Engineering Research Council 1991).
As geometry evolves to encompass the understanding of diverse visual phenomena, it is important to be clear about what is meant by the geometrical reasoning necessary to solve mathematical problems involving visual phenomena, and how such reasoning develops. The purpose of this paper is to provide an overview of several theoretical models which have been put forward as useful frameworks for describing and understanding the development of geometrical reasoning. After a brief outline of the van Hiele model of thinking in geometry, and of Fischbein's theory of figural concepts, a somewhat fuller description is provided of Duval's cognitive model of geometrical reasoning.
the van Hiele model of thinking in geometry One framework describing the development of geometrical reasoning that has been the subject of considerable research is the van Hiele model of thinking in geometry (see, for instance, van Hiele 1986). This is a teaching approach based on levels of thinking commonly known as the "van Hiele levels", originally aimed at the teaching and learning of geometry but which may be applicable more widely (Pegg 1992). In the van Hiele model there are at least 5 levels, although some writers discern as many as 8. The structure of the van Hiele model bears some similarity to the framework
Page 29
~ "
proposed by the UK Mathematical Association in 1923 which recognised three stages in the teaching and learning of geometry. These three stages were, briefly: stage A, intuitive, experimental work; stage B, 'locally' deductive work (formal sYmbolism and deductive reasoning is introduced, but intuition and induction are used to bridge logically difficult gaps); Stage C, globally rigorous work (Mathematical Association 1923). In a similar way, the van Hiele approach fosters the idea that students' initial curricular encounters with geometry should be of the intuitive, explanatory kind (van Hiele 1986 p 117). The learner then progresses through a series of 'levels' characterised by increasing abstraction. Fuys et al give the following description of the different levels, based on their translations of the work of van Hiele from the original Dutch: level 0 level 1 the student identifies, names, compares and operates on geometric figures the student analyses figures in terms of their components and relationships between components and discovers properties/rules empirically the student logically inter-relates previously discovered properties/rules by giving or following informal arguments the student proves theorems deductively and establishes interrelationships between networks of theorems the student establishes theorems in different postulational systems and analyses/compares these systems (Fuys et al1988 p5) The van Hie1e model has been subject to some critical discussion including querying, for example, the discreteness of the levels and the precise nature of levels 0 and 4 (or 1 and 5 as some writers denote them). For further details see Fuys et al1988 and, for reviews, Hershkowitz 1990 and Pegg 1992. the theory of figural concepts Fischbein (1993) observes that while a geometrical figure such as a square can be described as having intrinsically conceptual properties (in that it is controlled by a theory), it is not solely a concept, it is an image too. As he says" it possesses a property which usual concepts do not possess, namely it includes the mental representation of space property" (ibid p 141). So, Fischbein argues, all geometrical figures represent mental constructs which possess, simultaneously, conceptual and figural properties. According to this notion of figural concepts, geometrical reasoning is characterised by the interaction between these two aspects, the figural and the conceptual. Mariotti (1995 p94), in discussing Fischbein's notion of figural concept, stresses the dialectic relationship between a geometrical figure and a geometrical concept. She argues that geometry is a field in which it is necessary for images and
level 2 level 3 level 4
Page 30
concepts to interact, but that from the student's perspective there can be a tension between the two.
Duval's cognitive model of geometrical reasoning The French psychologist Duval approaches geometry from a cognitive and perceptual viewpoint. For example, in Duval (1995 p 145-147) he provides an analytic resource in the form of a detailed framework for analysing the semiotics of geometric drawings. In this framework he identifies four types of what he calls "cognitive apprehension". These are: 1. perceptual apprehension: this is what is recognised at first glance; perhaps, for instance, sub-figures which are not necessarily relevant to the construction of the geometrical figure. 2. sequential apprehension: this is used when constructing a figure or when describing its construction. In this case, the figural units depend not on perception but on mathematical and technical constraints (in the latter case this could be ruler and compasses, or perhaps the primitives in computer software). 3. discursive apprehension: perceptual recognition depends on discursive statements because mathematical properties represented in a drawing cannot be determined solely through perceptual apprehension, some must first be given through speech. 4. operative apprehension: this involves operating on the figure, either mentally or physically, which can give insight into the solution of a problem.
As Duval explains (ibid p 155), there is always a potential conflict between perceptual apprehension of a figure and mathematical perception: "difficulties in moving from perceived features of a figure can mislead students as to the mathematical properties and objects represented by a drawing, and can obstruct appreciation of the need for the discovery of proofs". According to Duval, operative apprehension does not work independently of the others, indeed discursive and perceptual apprehension can very often obscure operative apprehension. From a teaching perspective Duval argues for "special and separate learning of operative as well as of discusive and sequential apprehension are required". Duval suggests that work with computers may support not only the development of sequential apprehension, but also the development of operative apprehension, if the software has been designed with this in mind. HE concludes that "a mathematical way of looking at figures only results from co-ordination between separate processes of apprehension over a long time". While the above refers to working with geometric drawings, Duval (1998 p38-39) has gone further in proposing that geometrical reasoning involves three kinds of cognitive processes which fulfill specific epistemological functions. These cognitive processes are: visualisation processes, for example the visual representation of a geometrical statement, the or heuristic exploration of a complex geometrical situation . construction processes (using tools)
3\
Page 31
reasoning proces5,'es - particularly discursive processes for the extension of knowledge, for explanation, for proof Duval points out that these different processes can be perfonned separately. For example, visualisation does not necessarily depend on construction. Similarly, even if construction leads to visualisation, construction processes actually depend only on the connections between relevant mathematical properties and the constraints of the tools being used. Similarly, even ifvisualisation can be an aid to reasoning through, for instance, aiding the finding of a proof, in some cases visualisation can be misleading. However, Duval argues, "these three kinds of cognitive processes are closely connected and their synergy is cognitively necessary for proficiency in geometry" (ibid p38). Duval illustrates the connections between these three kinds of cognitive processes in the way represented in figure I below. In Figure I, each arrow represents the way one kind of cognitive process can support another kind in any geometrical activity. Duval makes arrow 2 dotted because, as argued above, visualisation does not always help reasoning. Arrows 5A and 58 illustrate that reasoning can develop in a way independent of construction or visualisation processes.
Given Duval's argument that the synergy of these three cognitive processes is cognitively necessary for proficiency in geometry, the issue is, as Duval identifies, how to get pupils in school to see the communication between these three kinds of processes. Duval argues that in attempting to understand the development of geometrical reasoning, his research shows the following: 1. The three kinds of processes must be developed separately. 2. Work on differentiating visualisation processes and between different reasoning processes is needed in the curriculum. 3. The co-ordination of these three kinds of processes can really occur only after this work on differentiation.
Page 32
Conclusions The above overview of three fairly well-developed frameworks for describing and understanding the development of geometrical reasoning is intended to provide a brief idea of the theoretical resources available which may be useful in research in this area. It also underlines the cognitive complexity of geometry. As Duval concludes (ibid pSI), much research about the deep process of the development and the learning of visualisation and reasoning are still needed". References Duval, R (1995), Geometrical Pictures: kinds of representation and specific processings. In R Sutherland and J Mason (Eds), Exploiting Mental Imagery with Computers in Mathematics Education. Berlin: Springer. Duval, R (1998), Geometry from a Cognitive Point of View. In C Mammana and V t Villani (Eds), Perspectives on the Teaching of Geometry for the 2r Century: an ICMI study. Dordrecht: Kluwer. Fischbein, E (1993), The Theory of Figural Concepts. Educational Studies in Mathematics. 24(2), 139-162. Fuys, D., Geddes, D., & Tischer, R. (1988), The van Hiele Model of Thinking in Geometry Among Adolescents. Reston, Va. National Council of teachers of Mathematics Hershkowitz, R. (1990). Psychological Aspects of Leaming Geometry. In P. Nesher & J. Kilpatrick (Eds.), Mathematics and Cognition. (pp. 70-95). Cambridge: CUP. Mariotti, M A (1995), Images and Concepts in Geometrical Reasoning. In R Sutherland and J Mason (Eds), Exploiting Mental Imagery with Computers in Mathematics Education. Berlin: Springer. National Research Council (1990), Renewing US Mathematics. Washington: National Academy Press. Pegg, J (1992), Students' Understanding of Geometry: theoretical perspectives. In: Southwell, B, Perry, B and Owens, K (eds), Space: the first andfinalfrontier, proceedings of the 15th conference of the Mathematics Education Research Group of Australasia. Sydney: MERGA. Science and Engineering Research Council (1991), Mathematics: strategy for the future. Swindon: SERC. van Hiele, P M (1986), Structure and Insight: a theory of mathematics education. Orlando, Fla: Academic Press. BSRLM Geometry Working Group The geometry working group focuses on the teaching and learning of geometrical ideas in its widest sense. The aim of the group is to share perspectives on a range of research questions which could become the basis for further collaborative work. Suggestions of topics for discussion are always welcome. The group is open to all.
33
Contact: Keith Jones, University of Southampton, Research and Graduate School of Education, Highfield, Southampton, SO 17 1 BJ, UK. e-mail: dkj@southampton.ac.uk tel: +44 (0) 1703 592449 fax: +44 (0)1703 593556 http://www.soton.ac. uk! ~gary/ crime.html
Page 34
ADVANCED MATHEMATICAL THINKING WORKING GROUP Convenor: Stephen Hegedus University of Southampton
ABSTRACT: Initially, this group will want to assess the interest in Advanced Mathematical Thinking (AMT), with the possible aim of creating links with other international working groups in the field (e.g. P ME). The group will be introduced to some of the major issues in AMT research today, and possible future developments. Themes of interest at the moment include: students' conceptions offunction, limits, proof, and students' conceptions of differentiation and integration. It will look at the relevance of social, pedagogical and methodological issues, such as the use of algebraic/geometric software or methods for analysing (meta-)cognitive behaviour, in developing a more suitable model of AMT. It is a major aim that the group will not concentrate on purely psychological issues but will discuss the nature of mathematics, per se.
The group met for the first time and a keen interest was shown by a dozen or so people in the field of Advanced Mathematical Thinking (AMT). The group was introduced to some of the main ideas of AMT and some of the main epistemological notIons which have been developed and criticised these past 18 years.
We initially discussed what AMT was and what was known so far. AMT was said to be a branch of the Psychology of Mathematics Education but fundamental questions about the nature of mathematical thinking and advanced problem-solving processes were asked. This was for the purposes of dealing with problematic issues in students' problems-solving process, epistemological obstacles (Sirepinska 1987) and pedagogy (Dubinsky & Tall, 1991).
The group was asked to think about possible discrepancies in a psychology of mathematics education and how we could address these issues. Many in the group were concerned with the word "advanced" and what it referred to; whether it was exclusive or elitist. We looked at parts of Tall (1991) which related advanced mathematical thinking to elementary mathematical thinking., and that the advanced did not always refer to advanced mathematics but advanced thinking. Analogies between elementary axiomatic learning and advanced mathematical
3
(under-) graduate thinking are represented in Tall (1991; Ch. 15) and Gray & Tall ( I 994), for example.
Primarily, we are concerned with advanced mathematics, since higher-order thinking is prominent in advanced mathematical problem-solving, but research in elementary thinking can highlight epistemological notions and ontological ideas in the former. How a student might conceptualise a function and manipulate functional notions; how a student might manipulate function in an ineffective way might well be to do with elementary functional work in axiomatic relationships.
*
The working group concentrated on two themes in the first session:
I. What has AMT offered to Mathematics Education? 2. How can we re-shape AMT (i.e. does the group have a purpose)?
Part I: What has AMT offered to Mathematics Education?
It is evident in a review of the literature (Hegedus, 1998) that AMT has offered us two things:
An epistemological framework of rigorous psycho-mathematical concepts which have addressed thinking in an advanced mathematical domain far more intuitively than some of the vague generic psychological terminology used in educational research to date. Data to increase our understanding of students' problem-solving behaviour. These have generally been incorporated into technological pedagogical packages.
The main epistemological notions which have been developed/discussed in the AMT literature are:
Page 36
1. Concept Image/Concept Definition. 2.Reflective Abstraction. 3. Procept.
The notions were explained to the group and examples were shown. The group responded with considerable interest to the final notion, the procept, and it was agreed to discuss this term in greater detail at the next meeting.
Examples form the Calculus (re Limits and function) and didactical engineering were extracted from Tall (1991; Chps. 9,10,11) and discussed. Some of the main epistemological obstacles discussed were: The failure to link geometry with numbers. The notion of the infinitely large with infinitely small. The metaphysical aspect of the notion of limit. Is the limit attained or not?
Part II: How can we re-shape AMT?
The group discussed the position of AMT as a research program and a school of psychological thought. The group adhered to ideas that elementary mathematical thinking could be examined in an AMT paradigm to relax the elitist nature that is inherent in the body of thought and which makes it possibly unapproachable for certain people.
The major theme was on acquiring an epistemological base with which we are confident with and with which we can discuss. From this base we might develop, deconstruct, or reconstruct epistemological notions in the AMT paradigm and work towards a more dynamic epistemology and suitable methodologies to create a better research environment.
The group believes that we need to examine what ideas have been put forward so far about the nature of Advanced Mathematical Thinking, to critically analyse them, and
3
to work towards a clearer understanding of the nature of advanced mathematical thinking.
I put forward three issues as momentum for the construction of a new psychology of AMT.
The existence of a dichotomy between traditional psychology and educational psychology. Psycho-linguistic vs. linguistics. Do we have some un-entangling to do? Generic psychological terms are too abundant. We need to analyse them and synthesise a formal psycho-mathematical discourse which is useful to the mathematics and mathematics educational community.
The group concluded with an aim to discuss the procept in further detail at the next meeting. Developments of the group and AMT resources are housed on the web at http://www.soton.ac. uk/~amt.
REFERENCES
Dubinsky, E., & TaU, D. O. (1991). Advanced mathematical thinking and the computer. In D. O. Tall (Ed.), Advanced mathematical thinking. The Netherlands: Kluwer Academic Publishers .. Gray, E. M., & TaU, D. O. (1994). Duality, ambiguity, and flexibility: A proceptual view of simple arithmetic. Journal for Research in Mathematics Education, 25 (2), 116-140. Hegedus, S.J. (1998). Analysing the metacognitive abilities of mathematics undergraduates solving problems in the integral Calculus. Doctoral Thesis presented at University of Southampton. http://www.soton.ac.ukl~heg/thesis/index.htm Sirepinska, A. (1987) Humanities Students and epistemological obstacles related to limits. Educational Studies in Mathematics, 18 (4), 371 - 387.
Page 38
RELATIONS BETWEEN TEACHER'S REPRESENTATIONS AND PUPIL'S IMAGES Chris Bills, University of Warwick
Abstract: This paper presents a simple constructivist model of teaching and learning characterised as a mapping between the set of representations that the teacher uses and the set of images that learners form. Data, collected in a phenomenographic study of primary school pupils' images, formed in consequence of their interaction with representations of two digit numbers, is discussed. Some implications for teaching are considered.
Introduction
My initial research interest was in the teaching and learning of fractions. Preliminary observations in Y2 and Y3 classes of a Solihull primary school, however, led me to the opinion that the difficulties that learners face in developing their conception of fraction is a particular instance of the general problems involved at each stage of expanding their concept of number. At any time when there is a need to introduce a larger set of numbers than pupils have previously experienced teachers engineer the encounter and pupils make sense of that encounter. The 'engineering' involves the teacher using a variety of representations to communicate to pupils the mathematics that exists in the teacher's mind. The 'sense' the pupils make of their interaction with these tangible embodiments involves forming a mental representation which they can use in thinking about the mathematics. The mental representation takes a variety of forms and is often confusingly called image (it need not be visual). Here the word image is used in the mathematical sense i.e. a transformation from the representation that has been given. This provides a model of teaching and learning:
It allows for the possibilities of one-to-one, one-to-many, and many-to-one mappings between these sets. In this paper I describe my research on the representations and images. I will start with an indication of my method then consider each component of the model separately. In conclusion I consider the use of the model to describe my research process.
Method
The data gathering strategies that I have developed during the preliminary observations period have been appropriate for my purpose of describing and analysing pupils' responses to their experiences of the representations. Field notes of the lessons record: Teacher's use of words, actions, writing, drawings, equipment and the tasks they set; Pupils' response to teachers' questions, their questions to the teacher, their behaviour in response to tasks set; Pupils' answers to my questions during the lesson on how they are thinking about the tasks set.
39
Semi-structured interviews with individual pupils are audio tape recorded and notes made on their physical activity which relate to their descriptions of images. Initial questions are typically on relations between, or operations on, numbers with follow-up exploratory questions of the form "How did you decide that?", "What was in your mind when you did that?", "How do you think about that sort of thing?". My model of teaching and learning is in the constructivist tradition, my assumption that pupils make sense of the mathematics by forming and using their own images is in the constructivist mould. My approach to the research is a naturalistic qualitative one which can be termed "phenomenographic" (Marton, 1988). Phenomenography is a research approach originally developed by researchers in the Department of Education of the University of Gothenburg. A phenomenographic study is an investigation of people's understanding of phenomena which seeks to categorise and explain the qualitatively different ways in which people think about the phenomena. The categorisation of descriptions is based on structurally distinctive characteristics of the responses in semi-structured interviews. The categories developed in one context are potentially part of a larger structure of categories applicable in other contexts. The initial discovery of previously unspecified categories of thinking may be peculiar to the researcher and context but the test of their validity is in their applicability for other researchers and as a source of explanation of differences in learning outcomes. I start with the assumption that the images formed by pupils can not be studied in isolation from the context of the classroom or the interaction between the pupils and the representations used by their teachers. It is essential that I observe, first hand, the common experiences of the learners as a basis for my analysis of their different conceptualisations. My focus is on what the teachers say and do and how the pupils respond to this by developing images based on the representations.
Representations
The term 'didactic transposition' (Kang & Kilpatrick, 1992) is used to indicate that the representations that teachers use are not the mathematics but a transformation of the mathematics into a communicable form. The types of representations that are available to teachers have been identified as: real world contexts; manipulatable models; pictorial; spoken language; written symbols (Lesh, Post, & Behr, 1987). In the lessons I have observed the representations used by the teachers tend to be influenced by the textbook writers. The teachers prepare the pupils to attempt exercises which are specific to a representation used in the books. I have focused on the teachers' attempts to extend the number sense of the children beyond their first experience of number as the easily countable cardinal number of a set. A variety of representations have been used: Objects to count into bundles of ten (cubes, straws, marbles in bags), Dienes arithmetic blocks, Verbal (counting on and back in powers of ten, chanting emphasising place value), Column notation (individual digit cards, numerals written in columns), Number line, number track, number square, Money, Spike Abacus. Two associated procedures are in evidence: NlO (counting on by tens - ordinal) and 1010 (decomposition in tens and units - cardinal) (Beishuizen, 1993)
Page 40
Field notes give a flavour of the explicit representation-specific language of the teacher, T. 17/10/97 Y2- Developing 1010 strategy using Dienes blocks: T gives Mandy 2 tens ('Another way of putting it ?' M says 'twenty') and 4 ones ('How many altogether ?' M says 'twenty-four'). Gives Nina 1 ten and 2 ones ('How many altogether ?' etc.). 'Now put them together in my hands' M and N put the tens in one hand and the ones in the other, 'How many altogether? Look how easy it is to add them instead of all individual cubes.' 27/2/98 Y2 - Some pupils demonstrate in a revision lesson that the separating of tens and ones has become part of their way of thinking: 'Now add 20p on. How will you do that?' Ann says 'see what the first number is and add 2 on to it' another pupil says 'add ten on then another ten', another says 'if you had 59 take the 9 off and add the tens on then add the 9 back'. The teacher reinforces 'what I don't want you to do is to do 39,40,41, like that, 'cos that would be silly.' Subsequently a variety of methods are in use for adding 20p on to different amounts of money including John counting on in ones using the number square and getting them wrong through miscounting.
Mapping from representation to image

The pupils listen, watch, and attempt the tasks set. From this activity they form their images. I follow Lesh and Kelly (Lesh & Kelly, 1997) in thinking that:
Humans interpret their experiences using internal conceptual structures, which cannot merely be received from others, but which must be developed, actively, by each individual. Further we assume that the meanings of these constructions tend to be partly embedded in a variety of external systems of representation, (p398)
They, like Mason (Mason, 1987) , reject the conveyance metaphor of teaching and learning i.e. that meaning is carried by representations and the learner must simply take out the meaning when the representation is received. If there is a hierarchy in the representations it is difficult to detect in the textbooks or from classroom observations. Thus it seems that these mathematically naive learners meet a heterarchical collection of representations and can 'actively develop' discrete images. Each representation has procedures which the teacher demonstrates to pupils whose subsequent successful performance of these representation-specific procedures can be misinterpreted as understanding of the underlying mathematics. It is possible for the pupils to adopt each representation-specific procedure separately without recognising the common mathematics i.e. without developing their concept. They add tens by: saying next word in word sequence; going down a column on a number square; taking a ten-step on a number line; replacing tens digit with one higher; adding one to the tens column; having an extra ten-block; having an extra lOp coin. Many find one representation easier to use than others so use it exclusively and cannot explain one representation's procedures in terms of another. The representations and their procedures are sufficiently different in appearance to seem like different concepts. The teachers observed have provided representations in the attempt to enhance previous images and have helped pupils re-collect the experiences that have given rise to them, in order to develop more efficient procedures. However the shared image that was the teacher's goal was not necessary for effective completing of exercises. Many pupils were reluctant to abandon previous dominant
44
images (e.g. addition by counting-on in ones) which had served well in the past. It could be that there is less expenditure of mental energy needed to use a slow previous procedure than to drive the imagination required to form a more robust image. The desire for efficiency, whether due to enculturation or instinct, may determine that pupils stay with the tried and tested rather than invest the effort required to change. The frequent recall of previous images to aid the formation of new ones could also have the effect of strengthening them. It is apparent that there is not a one-to-one mapping from the teacher's image to pupils' images mediated by representations. The teacher may choose a variety of representations that embody the same mathematics but each medium can become a separate message.
Images
(Russell, 1956) gives a definition:
An image is a centrally aroused experience which reproduces in part some previous perceptual experience in the absence of the original sensations. (p68)
where 'perceptual' is taken as a reaction to any stimulus i.e. not only visual. The images, though, are likely to reflect the experiences that give rise to them (Lawler, 1996). It is also suggested that a learner's "concept image", consisting of mental pictures, properties and processes, may develop with each new experience (Tall & Vinner, 1982). I have used the word 'image' to mean that which the learner makes use of in his thinking as a result of his interaction with the representation. In interviews pupils may not mention a teacher's representation but it could still have left a 'trace' (akin to construction lines, or the trace left after an object has passed, or the trace element that is so watered down it is almost undetectable) which influences the development of their image. The image developed by learners as a result of exposure to representations are more-or-less accurate re-collections of their experiences transformed by their imaginations. I take 'imaginations' in this sense to mean the manipulation of images into a form which the learner finds comprehensible. The image formed may lose many of the subtleties of the original representation as it is moulded to fit with the learners dominant preVIOUS Images. In pilot semi-structured interviews I have encouraged pupils to describe what they are thinking when they add 53 to 24, how they decide what comes after 276, how they work out 86p-12p. I have explored a variety of categorisations : Recency - Some pupils make explicit reference to a recently introduced representation and associated procedures, others use strategies they have developed themselves, others call up their most dominant previous image; Modality - Some pupils describe "seeing" blocks or squares or numerals, others "hear" words in their own or their teacher's voice; Variety - Some pupils use the same representation for each question others use a different representations for each; Progression - The majority of pupils interviewed early in Y2 used counting on and back in ones whilst Y3 pupils predominantly separated tens and ones and made reference to written algorithm procedures; Completeness - Some pupils made reference to a representation but could not recall the procedures; Flexibility - Few pupils could use more than one image for a particular question.
Page 42
Exerts from transcripts of interviews with Y2 pupils conducted in October when they had been using both number squares and Dienes blocks give an indication of the range of images: 1) Hazel has developed a strategy that is neither 1010 nor NlO and she makes no reference to any of the representations that have been used in class recently. My assumption that she would use one of these means that my questioning and prompting are representation-specific. The first question, which was designed to see if she would use a taught procedure, was interpreted as a request for her to calculate in a way she chose, perhaps because she had no image to support the suggested procedures. The fact that I failed to persuade her to use an image based on one of these representations could indicate that she has not formed one because her own is sufficient for her needs. She has not incorporated the teacher's representation-specific procedures into her image because they do not fit with it. I Do you remember how to do those? (points at sum, 24 + 53, printed on paper) H gave answer 77 after moving fingers for 27 seconds I When you added those you were counting with your fingers weren't you. Can you tell me what you were counting? H I counted 53 and 24. I Good so you started with the big number and then you added 24. Did you add 20 first or 4 first or did you count on 24? H I added 5, 5, 5, 5, and then the 4. I Would you have liked to use the 100 square or the rods and blocks to do that or did you like doing it like that in your head? H I like doing like that. I Can you write the answer down for me? H Forgotten the answer. H starts counting again saying the numbers semi-audibly, using just one hand, counting on five at a time remembering how many fives she had added and got the same answer in 20 sees. 2 Neal has an image of 2-digit addition that uses the procedure requiring the digits to be separated but does not have the procedure which adds tens to tens and units to units I If you had to do this in your head what would you do? N You get 24 in your head then add on the 5 and the 3. I So what would that give you then? N 32. ( 8 sees) 3 Nina has some recall of the procedures related to the Dienes blocks but the overload caused by attempting to manipulate this image means that she can not make use of it: I Last week you were doing these (points to sum). You were doing them with rods and blocks and you had been doing them with a hundred square but if you had to do that in your head could you do it? Would you like to try? N ... Don't know. (the answer) I What would you start with? N 24 .. and then you would add 53. I Would you like to try it N ... Don't know. (the answer) I If I had brought the rods and blocks and the hundred square which would you choose? N Blocks I You like using those best. If you had the blocks what would you do with them. N I would put 24 and 5 big ones. I And then you would be able to put them together. Can you imagine that? If there were 24 and 5 big ones can you work that out? N ... don't know.
4-
Page 43
~
Implications for teaching and learning If learners are to develop their concept it is insufficient simply to have the ability to recall representation-specific procedures in response to representation-specific questions. I suggest that four steps are required: Re-mind, Re-collect, Re-cognise and Real-ise. In order to make a new representation meaningful (related to other representations) the teacher has to re-mind pupils of the other representations - she encourages them to bring back to their minds those previous encounters with this mathematics. Pupils need to re-collect those images that they have previously formed then to re-cognise that their new image is related to the others. The act of re-cognition implies a sense of new understanding of this and previous images in the light of the links between them. Finally real-ise signifies the act of making real. A learner only realises what mathematics is conveyed by a representation when they have an image of it that is real for them. Conclusion As a naive/novice learner in this field I am attempting to build images from perceptions and representations of the learning process. Since I can only understand when I can attach an image to other images that I already have I may resort to analogies to give me an opportunity to recognise. In order to communicate part of my understanding of teaching and learning I give this representation which I refer to as "Representations and Images" using words that symbolise my image. In this act of didactic transposition I inevitably transform the image and again the medium becomes the message. The purpose of the study is to present a representation of teaching and learning that might provide an image for other teachers to use. References Beishuizen, M. (1993). Mental strategies and materials or models for addition and subtraction up to 100 in Dutch second grades. Journalfor Research In Mathematics, 24(4),294-323. Kang, W., & Kilpatrick, J. (1992). Didactic Transposition in Mathematics Textbooks. For the Learning of Mathematics, 12(1), 2-7. Lawler, RW. (1996). Thinkable Models. Journal of Mathematical Behavior, 15(3),241-259. Lesh, R, & Kelly, E. (1997). Teachers' evolving conceptions of one-to-one tutoring: A threetiered teaching experiment. Journalfor Research In Mathematics, 28(4),398-430. Lesh, R, Post, T., & Behr, M. (1987). Representation and Translations among Representations in Mathematics Learning and Problem Solving. In C. Janvier (Ed.), Problems of representation in the teaching and learning of mathematics (pp. 33-40). Hillsdale,NJ: Erlbaum. Marton, F. (1988). Phenomenography: A Research Approach to Investigating Different Understandings of Reality. In R Sherman & RB. Rodman (Eds.), Qualitative Research in Education: Focus and Methods. Lewes: Falmer Press. Mason, J. (1987). What do symbols represent? In C. Janvier (Ed.), Problems of representation in the teaching and learning of mathematics (pp. 73-81). Hillsdale,NJ: Erlbaum. Russell, D.H. (1956). Children 's Thinking. Boston: Ginn. Tall, D.O., & Vinner, S. (1982). Concept image and concept definition in mathematics with particular reference to limits and continuity. Educational Studies in Mathematics, 12, 151-169.
Page 44
VISUALISATION AND USING TECHNOLOGY IN A LEVEL MATHEMATICS Sally Elliott, Sheffield Hallam University This project seeks to identify and evaluate the ways in which existing technology can be utilised to promote and develop students' powers of visual is at ion and to encourage the usage of these skills in A level mathematics lessons. At present, a small scale pilot study has been carried out and the resulting data has been analysed. After briefly summa rising current research in the area of visual is at ion and technology in this paper, I will report on the findings of this initial pilot study, in which materials developed for use with the TI-92, aimed at promoting students' abilities to visualise the graphs of functions, were trialled with a class of thirteen year twelve students. The subsequent consequences for future directions of the research will also be discussed. Introduction Visualisation is increasingly being accepted as an important aspect of mathematical reasoning. Studies have revealed that 'activities encouraging the construction of images can greatly enhance mathematics learning' (Wheatley and Brown, 1994). Indeed, potentially, technology could assume a very powerful and influential role in stimulating and shaping students' powers of visual is at ion, and as such may prove to contribute significantly to the depth of students' understanding.
Zimmerman and Cunningham (1991) insist that mathematical visua1isation is not merely 'math appreciation through pictures' - a superficial substitute for understanding. Rather they maintain that visualisation supplies depth and meaning to understanding, serving as a reliable guide to problem solving, and inspiring creative discoveries. In order to achieve this understanding, however, they propose that visualisation cannot be isolated from the rest of mathematics, implying that symbolical, numerical and visual representations of ideas must be formulated and connected. This project is conceptualised on the basis that visual thinking and graphical representation must be linked to other modes of mathematical thinking and other forms of representation (Tall, 1989). Issues Surrounding Visualisation Within the current literature there exist many differing notions of the key terms associated with the area of visualisation in the learning of mathematics, each developed with respect to a specific research purpose/focus, and each drawing on and expanding previous ideas.
Mariottii and Pesci (1994) acknowledge visualisation occurring when 'thinking is spontaneously accompanied and supported by images'. Mason (1992) regards visualising as 'making the unseen visible' and imagery as 'the power to imagine the possible and the impossible'. Solano and Presmeg (1995) see visualisation as 'the relationship between images' - 'in order to visualise there is a need to create many images to construct relationships that will facilitate visualisation and reasoning'. Hitt Espinosa (1997) suggests that visualisation of mathematical concepts is 'not a trivial cognitive activity: to visualise is not the same as to see'. To visualise is the 'ability to create rich, mental images which the individual can manipulate in his mind, rehearse different representations of the concept and, if necessary, use paper or a computer screen to express the idea in question'.
4--5 Page 45
Unfortunately, despite the current views of researchers surrounding the importance of visualisation, there is still a tendency for visualisation to be undervalued in mathematics classrooms and consequently some students, whilst able to visualise mathematically, often opt for non-visual, more 'conventional' approaches to problem solving (Presmeg, 1995). Traditionally, a greater emphasis has been placed on algebraic or analytic proof, despite the proposed legitimacy of visual theorems. Presmeg's findings (1986) indicate that an ability to apply and interchange both visual and non-visual methods in problem solving is particularly advantageous for students, especially where one mode is more appropriate. However, the teaching of school mathematics is predominately non-visual and 'visualisers are seriously under-represented amongst high mathematical achievers' (ibid). Although, images presented to students by teachers will influence the students' understanding and individual construction of such images, the students' conception of these images will not necessarily correspond to that of the teachers' (Mason, 1992). Indeed, 'visual ideas often considered intuitive by an experienced mathematician are not necessarily intuitive to an inexperienced student' (Tall, 1991). Students should be encouraged to create and explore their own images (Cunningham 1994) - a visual understanding of a given situation is more robust and is thus more likely to be remembered by the student in the longer term than a purely algebraic proof. Yet, Presmeg (1986) outlines four particular difficulties involving imagery; images/diagrams viewed inappropriately, inflexible thinking when dealing with a non-standard diagram, rigid uncontrollable images and vague imagery. She (ibid), also, suggests that 'less imagery is used with greater experience or learning'. Visualisation skills may be employed by students privately to clarify, interpret and make sense of the given problem intuitively, as tools for 'meaning-making' (Wheatley and Brown, 1994) although, such processes are unlikely to be explicit in written arguments (Presmeg, 1995). Furthermore, the usage of visual techniques is comparatively time intensive suggesting that tests and examinations will tend to favour the non-visual thinker (Presmeg, 1986). In addition, visual thinking requires nonsequential, parallel processing of information, and as such poses a greater cognitive challenge to students than step by step sequential algorithmic reasoning (Eisenberg and Dreyfus, 1991). The Perceived Role of Technology In light of the recent advancements in technology, a whole range of computer programmes and scientific instruments are currently available with the potential to assist students in the formation of visual mathematical images. One of the main objectives of this research is to evaluate and develop materials and strategies which aim, as far as possible, to maximise this potential, with particular emphasis on the graphical calculator. However, the role of the computer in this respect is, also, regarded by the researcher as extremely important and influential and is thus explored in this review of current literature. Overall, the findings of studies involving graphical calculators appear to be very similar to those which utilised computer technology, although the similarities and differences between these two types of technologies should not be overlooked.
Page 46
Many researchers realise the potential of utilising technology to promote and encourage visualisation skills (Souza and Borba, 1995; Smart, 1995). In particular, computer based visual approaches in teaching mathematics can i) increase motivation and ii) provide an opportunity to pursue an alternative and yet complimentary mode of thought to the traditional symbolic approach (Cunningham, 1994). Technology can be utilised to enable students to develop a deeper insight into the relationship between functions and their graphs (Carulla and Gomez, 1997). Furthermore, technology can be particularly useful in exploratory learning, where students are able to formulate concepts for themselves and benefit from visualisation in the process (Tall, 1991). However, despite the advantages, students may still misunderstand, misinterpret and therefore misuse information provided by graphic calculators (Carulla and Gomez, 1997). Furthermore, such technology may encourage students to focus primarily on graphical representation whilst neglecting other modes (ibid). In contrast, other researchers report that graphical calculators can be utilised to foster the transitions between and exploration of different modes of representation (Ruthven, 1990). Multirepresentational software, however, could contribute towards misunderstanding and confusion amongst students; any difficulty experienced with one particular representation could be intensified by the presence of other forms of representations (O'Reilly et al,1997). There is a danger that students could become 'saturated by images' (Mason, 1992). Alternatively, students may become too dependent on technology, regarding the solutions generated as irrefutable (Smart, 1995). Zimmerman and Cunningham (1991) believe that certain fundamental visualisation skills are prerequisite for meaningful computer based visualisation. Many researchers maintain that the use of technology can promote collaborative learning and equal opportunities (Smart, 1995). In particular, female students, have benefited from the private nature of the graphics calculator (ibid). Ruthven (1990), also, found that a reduction in student uncertainty and anxiety accompanied regular use of the graphics calculator, and hence stimulated improvement in the 'confidence, competence and performance' of all students, especially that of the females. This study will investigate how graphical calculators affect the visualisation capabilities of the female students in comparison with the males, with the aim of determining whether female students benefit, in this manner, to a greater or lesser extent. The First Pilot Study Initial classroom trials were carried out at a school in Sheffield for a period of six hours during February 1998, with a group of five male and eight female year twelve students. The fieldwork involved participant observation and a post-trial questionnaire. Each individual student was given a TI92, although they generally worked together in pairs, sharing ideas. The exercises featured graphing functions, and involved exploring and identifying the effects of transformations, finding inverse functions, solving equations - graphically and algebraically, and investigating trigonometric and logarithmic identities. The main aim of this pilot study was to enable the researcher to assess the suitability of early materials and techniques, to elicit preliminary reactions to the use of technology and to provide a framework for further data collection.
4-1 Page 47
Student Questionnaire Responses The questionnaire responses indicated that, generally, this particular group of students viewed technology as an important addition to the A level mathematics classroom - a quick and accurate means of strengthening their understanding and visualisations of functions. However, some feared overdependency and accompanying laziness, but would nevertheless welcome further use of technology in the future. Thus, these students appeared to appreciate the opportunity to use the TI-92 and seem to have benefited mathematically from the experience. The Students' Work A preliminary examination of the students' work revealed that a high proportion of students often assumed that the TI-92 was displaying the whole graph without using the zoom in and out facilities. The function x2- x3 caused particular problems. The students were asked to sketch the graph of the function and to determine the nature and co-ordinates of any turning points. The first graph (fig I) is drawn using ZoomStd (where the x and y axes vary from -10 to 10, in divisions of I unit), and the second graph (fig 2) results from zooming in on the first to a degree of factor six, centred on the origin. The second graph provides a much better picture of the actual shape of the graph. Yet, all of the students who attempted this question failed to use the zoom facilities and thus drew a sketch of the function which resembled fig 1. Consequently, they mistook the point (0,0) as a point of inflection (clearly a local minimum in fig 2) and were unaware that a local maximum existed and as no-one checked their results by differentiation these errors were undetected. Clearly, these turning points were missed because they were not initially visible on screen.
In contrast, some students believed that the graph of y = (x+ 1)/(x+2)2 had a minimum turning point at x
= -2. These students failed to realise that the function is undefined at this x value as they completely
misinterpreted the graphs displayed by the TI-92 and neglected to inspect the equation.
Page 48
Fig 3 shows the graph of y = (x+ 1 )/(x+2)2, using ZoomStd again, whereas fig 4 is obtained by zooming out on the original graph, centred on the origin, by a factor of three. Since only part of the graph appeared to be visible on the screen in ZoomStd, some students choose to zoom out, producing graphs resembling fig 4, which seemed to have a minimum stationary point, and so these particular students (who were using the zoom facilities) were fooled. As before, these students did not spend time thinking logically about the function or picturing what the function might look like for themselves - they were confident that the technology provided them with the correct answer. Smart's (1995) research emphasises this problem, referred to as the 'magic:' element of technology.
In addition, few students successfully completed the algebraic components of certain questions, and fewer still actually specified the symbolic form of the graphs resulting from a series of successive transformations, even though this was requested. Thus, these students tended to concentrate on graphical representation in questions involving both graphical and algebraic aspects. However, the remaining questions were completed satisfactorily. In particular, eleven students were able to identify the graphs of all six functions in the final exercise, which was an encouraging outcome. Implications for Future Data Collection The data collected in this pilot study has enabled the first evaluation of the classroom materials and approaches devised by the researcher, aimed at promoting the development of student's powers of visualisation using technology to be undertaken, thereby permitting some initial progress in terms of achieving the second objective of the research. However, there was insufficient data to provide notable insight into the third and fourth objectives; to investigate the ways in which the technology acts as a tool in mediating the development of students' powers of visualisation and to investigate how powers of visualisation might be evoked and be developed by the use of mathematical software. To what extent did the materials encourage, if at all, visual thinking?
Thus, preliminary results suggest that it would be useful to establish a means which would indicate how students initially approach problems involving functions. In other words, before the research takes place, do they adopt a predominately visual, algebraic or numeric approach? If their approach tends to be visual, how successful are they? If their approach is not visual, how do they perform when asked to work visually? Moreover, following the introduction of technology does their preferred mode of operation change? Do the visualisation skills of all students (not only those who prefer to work visually) improve? Do all students necessarily have a visual approach? Teachers will be interviewed to determine the extent to which they have used visual methods in their teaching of functions, and in lessons generally. It is, also, recognised that to try to distinguish between visualisers and non-visualisers is problematic; there is a continuum between students who can be regarded as almost entirely visual thinkers and those who are virtually exclusively non-visual and furthermore there are few students at the extremes - for different types of problems individual students may use different methods of solution. In the future pre-trial exercises, questionnaires and interviews will be utilised in an attempt to ascertain answers to these central research questions.
Page 49
References Borba, M. (1996) Graphing Calculators, Functions and Reorganisation of the Classroom. Proceedings of Working Group16, leME-8, pp. 53-60. Carulla, C. & Gomez, P. (1997) Graphics Calculators and Problem Solving. Do They Help? Proceedings of PME 21,1, P 224. Cunningham, S (1994) Some Strategies for Using Visualisation in Mathematics Teaching. Zentralblattfur Didaktik der Mathematik, ZDM, 94/3, pp. 83-85. Eisenberg, T. & Dreyfus, T. (1991) On the Reluctance to Visualize in Mathematics. In Zimmerman, W. & Cunningham, S. (Eds.), Visualization in Teaching & Learning Mathematics, pp 25-37. Hitt Espinosa, F. (1997) Researching a Problem of Convergence with Mathematica: History and Visualisation of a Mathematical Idea. International Journal of Mathematical Education in Science and Technology, 28, N05, pp. 697-706. Mariotti, M. A. and Pesci. A. (1994) Visualization in Teaching - Learning Situations. Proceedings of PME 18, 1, P 22. Mason, J. (1992) Towards a Research Programme for Mental Imagery. Proceedings of the November Conference of BSRLM, pp. 24-29. O'Reilly, D. et al (1997) Constructive and Instructive Representation. Journal of Information Technology for Teacher Education, 6, No 1, pp 73-92. Presmeg, N. (1995) Preference for Visual Methods: An International Study. Proceedings of PME 19, 3, pp.58-65. Presmeg, N. (1986) Visualisation and Mathematical Giftedness. Educational Studies in Mathematics, 17, pp. 297-311. Ruthven, K. (1990) The Influence of Graphic Calculator Use on Translation from Graphic to Symbolic Forms. Educational Studies in Mathematics, 21, pp. 431-450. Smart, T. (1995) Visualisation, Confidence and Magic: The Role of Graphic Calculators. In Burton, L. & Jaworski, B. (Eds), Technology in Mathematics Teaching, pp 195-212. Solano, A. & Presmeg, N. C. (1995) Visualisation as a Relation of Images. Proceedings of PME 19, 3, pp. 66-73. Souza, T. & Borba, M. C. (1995) Quadratic Functions and Graphing Calculator. Proceedings of PME 19, 1, P 253. Tall, D. (1991) Intuition and Rigour: The Role of Visualization in the Calculus. In Zimmerman, W. & Cunningham, S. (Eds.), Visualization in Teaching & Learning Mathematics, pp 105-119. Tall, D. (1989) Concept Images, Generic Organizers, Computers, and Curriculum Change. For the Learning of Mathematics, 9, No 3, November, pp. 37-42. Wheatley, G. & Brown, D. (1994) The Construction and Re-presentation of Images in Mathematical Activity. Proceedings of PME 18, 1, P 81. Zimmerman, W. & Cunningham, S. (1991) What is Mathematical Visualisation? In Zimmerman, W. & Cunningham, S. (Eds), Visualisation in Teaching and Learning Mathematics, pp. 1-9.
50
What can we all say?" Dynamic geometry in a whole-class zone of proximal development.
John Gardiner Sheffield Hallam University 106506.2150@compuserve.com This paper first develops a theoretical background which culminates in a discussion of the promotion and fostering of socio-mathematical norms (Cobb and Yackel, 1996), the significance of local communities of practice (Winboume and Watson, 1998) and the development of a whole-class ZP D (H edegaard, 1991, Lerman, 1998). It then seeks to indicate a way in which classroom aproaches by teacher and pupils to a dynamic geometry package (Cabri 2 as available on the TI 92) might use this background Theoretical Background Vygotsky(1962) proposed a social background to learning and formulated the Genetic Law of Cultural Development, proposing that learning moves from the social to the personal. He took up the idea of the Zone of Proximal Development as the area where interaction between the individual and the social leads to development. Lerman (1998) has pointed out that his original definition was in terms of intelligence testing and the idea of the ZPD has been developed by later workers. The broadening of the definition over time is perhaps an indication of the wide applicability of the concept. It has been used both as a precise analytical tool and as a useful broad band of ideas. Definitions of the ZPD have ranged from those related to development and 'scaffolding', (Vygotsky, 1962, Bruner, 1985 :25), to assessment (Vygotsky 1956 :446-448), to the distinction and interaction between spontaneous and scientific concepts, (Kozulin, 1986, Gardiner and Hudson 1998) to definitions in activity theory (Engestrom 1987, 174), and to the class and teacher as a whole Hedegaard (1990). Lave and Wenger (1991 :48-49) discuss this movement to a broader cultural definition of the ZPD.
Of particular interest here is a definition of the ZPD which includes the classroom as a whole, in this case incorporating the teacher, the pupils and the technology. Hedegaard has reported (1990) in terms of the development of a whole-class ZPD rather than the analysis of an individual's learning. "This activity, in principle, is designed to develop a zone of proximal development for the class as a whole, where each child acquires personal knowledge through the activities shared between the teacher and the children and among the children themselves." (p 361). Hedegaard reports in the same paper a motivational shift in children's focus, from an interest in the concrete to interest in the derivation of
Page 51
~\
principles which can be applied to the concrete. Lerman (1998) takes the discussion further. " The ZPD is the classroom's, not the child's. In another sense it is the researcher's: it is the tool for analysis of the learning interactions in the classroom (and elsewhere). "(p 71)
Lave and Wenger (1991) see learning as situated in society and introduce the concept of 'Legitimate Peripheral Participation'. They bring forward the idea of masters and apprentices( or newcomers) and emphasise that situated learning is often independent of teaching. In discussion of reports of five master/apprentice situations they say: "In all five cases ..........researchers insist that there is very little observable teaching; the more basic phenomenon is learning." (p 92) Drawing on further work by Lave (1993), Winboume and Watson (1998) have used the idea of 'local communities of (mathematical) practice'. They provide a useful definition of such a community of practice and a summary of the features which are necessarily present in a classroom if those within it 'are to constitute a local community of practice: 1. Pupils see themselves as functioning mathematically within the lesson; 2. Within the lesson there is public recognition of competence; 3. Learners see themselves as working together towards the achievement of a common understanding; 4. There are shared ways of behaving, language, habits, values and tool-use; 5. The shape of the lesson is dependent upon the active participation of the students; 6. Learners and teachers see themselves as engaged in the same activity. '(p 183) This approach is echoed in the work of Cobb and Yackel (1996), who have analysed mathematics classrooms in terms of the negotiation and maintenance of social and sociomathematical norms. Social norms include insistence on explanation of answers, respecting the contribution of others and making clear agreement as well as disagreement. Socio-mathematical norms would include some notion of what constitutes a valid, complete solution, and agreement on the worth of alternative solutions. Negotiation must take place on how teacher and students agree on the mutual acceptability of solutions. Of course social norms will exist in all classrooms, and will bear a direct relationship to the society in which the classroom is situated. Because social norms will affect the negotiation of socio-mathematical norms, Apple (1992) has argued that the classroom must be frrmly situated in the wider context of the practices of school and society. I wish to argue here that these approaches, of a whole class ZPD, of a recognition of local communities of practice, and of negotiated socio-mathematical norms have much to offer in looking at how technology can be used in the classroom. In particular I want to indicate how I have tried to apply these ideas in the development of classroom material and techniques for use with the TI 92 hand-held computer in lower school teaching.
Page 52
The material The research discussed here took place in a rural comprehensive school in the UK. The main body of the research was a series of lessons taught by the researcher with one particular class and this will be reported on in a further paper. In addition the researcher was asked to provide most classes in year 9 (age 14-15) with a brief (two one hour lessons) introduction to staff and students of the possibilities of dynamic geometry. This requirement for a short course and for applicability to all ability ranges in the school led to the thinking behind the present paper. Each student had a TI 92 and an OHP version was available for demonstration by pupils and the teacher to the whole class. The following examples were an attempt to set up possibilities for whole class meaning-making with the minimum of previous knowledge of the TI92. The pattern followed was for the class to generate and discuss a simple dynamic image, and to record the result in exercise books as a diagram after the dynamic image had been appreciated. The hand-held nature of the TI92 is particularly suitable for pair discussion, and indeed, for consigning to a comer of the desk when work on paper is preferred. The exercises 1. The class were asked to draw a circle and a triangle with its vertices on the circle, then to measure the area of the triangle. They were then asked to investigate the effect of dragging one of the vertices.
Questions from the teacher/researcher were first directed at making sure that the class knew that there was no 'right' answer for the maximum area of the triangle. They were then asked to discuss whether there was any conclusion which could be drawn which was common to all their displays when the area of the triangle was a maximum. "What can we all say?" was a question used to unite the class and to try to draw them all into a local community of practice.
Page 53
Figures 2 and 3 illustrate another exercise, again using the minimum of previous knowledge, which seemed to the researcher to generate an opportunity for productive whole class meaning-making. Pupils were asked to draw a triangle (using the Triangle command) and then reflect it in one side (Fig 2) or rotate it through 180 degrees about the mid-point of one side (Fig3) and investigate the quadrilaterals (or triangles in the case of Fig 2) which could be produced by dragging the comers of the original triangle. Again the question "What can we all say?" was used to try to draw the whole class together, before going on to questions such as "How many rhombuses can there be?"
Page 54
Another exercisel which is available after only the briefest of introductions to the technology is based on a diagram such as Figure 4. Here pupils were asked to draw a triangle with its base on one of a pair of parallels and the vertex on the other. They were asked to investigate the affect on the area of the triangle of moving its defining points.
Conclusion Lave and Wenger (1991, pl02,103) address the issue of the transparency ofa resource, and this is further examined by Adler (1998, pp8-11). A resource used in a mathematics classroom can be so visible to students that it obscures the mathematics and prevents meaning making. At the same time some visibility is necessary. We want the resource to be visible in the sense that it should direct the gaze of students, so enabling their meaningmaking. 'Invisibility of mediating technologies is necessary for allowing focus on, and thus supporting the visibility of, the subject matter. Conversely, visibility of the significance of the technology is necessary for allowing its unproblematic- invisible use. This interplay of conflict and synergy is central to all aspects of learning in practice: it makes the design of supportive artifacts a matter of providing a good balance between these two interacting requirements.' (Lave and Wenger, 1991 pl03) Clearly the familiarity of students with technology such as the TI 92 governs its use, in a way which is informed by arguments such as this. As they become more familiar with the software the teacher will be able to introduce the use of more complicated functions without losing transparency. In the present case however, with students who had not used the machines before, the same arguments led to the approach outlined above. Questioning such as that outlined above is the stock in trade of many classroom practitioners, but consideration of the issues raised here has helped me to examine why some of my classroom techniques might work, and how I might look for routes to improvement.
References Adler, J (1998) Resources as a verb: Recoceptualising resources in and for school mathematics. in Olivier, A and Newstead, K, Eds,PME 22: International Group for the Psychology of Mathematics Education Conference, 1, 1-18 University of Stell en bosch, South Africa, July 1998. Apple M W (1992) Do the standards go far enough? Power, policy and practice in mathematics education. JRME 23,412-431 ' Bruner JS (1985) Vygotsky: a historical and conceptual perspective. In
I This example owes much to the specimen lesson from a Japanese school discussed, with the subsequent whole-class investigation, in the TIMMS video.( Stigler and Hiebert, 1997) An OHP dynamic image is presented to the class, and the pupils' work is done in exercise books, pointing to an alternative, and perhaps less expensive, approach.
Page 55
5 5
Wertsch JV, (ed) Culture, Communication and Cognition: Vygotskian Perspectives. Cambridge, CUP pp21-34 Cobb P and Yackel E (1996) Constructivist, emergent and sociocultural perspectives in the context of developmental research. Educational Psychologist 31,3-4, 175-190 Engestrom Y, (1987) Learning by expanding. Helsinki, Orienta-Konsultit Oy Gardiner, J. and Hudson, B. (1998) The evolution of pupils' ideas of construction and proof using hand-held dynamic geometry technology, in Olivier, A and Newstead, K, Eds,PME 22: International Groupfor the Psychology of Mathematics Education Conference, 2, 337-344 University of Stellenbosch, South Africa, July 1998 Hedegaard M (1990) The zone of proximal development as basis for instruction, in L C Moll (ed) (1990) Vygotsky and Education, Cambridge, Cambridge University Press, pp. 349-71 Kozulin A (1986) The concept of activity in Soviet psychology: Vygotsky, his disciples and critics, American Psychologist, 41 (3) 264-74 Lave J (1993) The Practice of Learning, in Chaiklin S & Lave J (eds) Understanding Practice: perspectives on Activity and Context, New York Cambridge University Press, pp 3-32 Lave J and Wenger E (1991) Situated Learning: Legitimate Peripheral Participation, Cambridge, CUP Lerman S (1998) A Moment in the Zoom ofa Lens: Towards a Discursive Psychology of Mathematics Teaching and Learning. in Olivier, A and Newstead, K, Eds'pME 22: International Group for the Psychology of Mathematics Education Conference, 1, 66-81 University of Stellenbosch, South Africa, July 1998 Stigler J W and Hiebert, J (1997) Understanding and Improving Classroom Mathematics Instruction: an Overview of the TIMMS Video Study, Phi Delta Kappan, Sept 1997. Vygotsky L S (1956) lzbannye psikhologicheskie issledovaniya [Selected Psychological Investigations] Moscow, Izdatel'stvo Akedemii Pedagogicheskikh N auk Vygotsky L S (1962) Thought and language, Cambridge, Mass, MIT Press Winbourne P and Watson A (1998) Participating in Learning Mathematics Through shared Local Practices, in Olivier, A and Newstead, K, Eds,PME 22: International Group for the Psychology of Mathematics Education Conference, 4, 177-184 University of Stellenbosch, South Africa, July 1998
510 From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 56
Cabri as a cognitive tool - Part 2

Bibi Lins Graduate School of Education University of Bristol Process of meaning production Amongst others, Bottino and Furinghetti (1994) investigated the ways in which teachers' belief are related with respect to two domains seen as correlate: Mathematics and Computers. My study, however, instead of examing the perceptions in two domains, I shall concentrate on how a new domain is constitued, namely the use ofICT in mathematical education. The proposed study takes place inside what Cooney (1994) calls a 'conception of teacher education as a work of investigation', precisely by suggesting and investigating the constitution of a domain.
Research Focus As I am doing research on ICT in school mathematics I could take any general/educational software that has been used for teaching mathematics. For instance, spreadsheets for teaching algebra or Geometer's Sketchpad for teaching geometry. Here I will facus on Cabri-Geometre, a dynamic geometry software. The approach taken views the designer (group) of Cabri as author, Cabri as text and the mathematics teacher as reader, with meaning production covering this complexity. Approach and Theoretical Framework Steve W oolgar et al. (1997) view technology as text via exploring of a metaphor: the machine as text. The frame of software as text is being explore here. The theoretical framework taken here is related to the epistemological model developed by Lins (1992). Meaning production involves at least three elements: author, text and reader. These elements are widely perceived as functioning together to produce communication in the sense of meaning being conveyed from author to reader via text. I will be working with an alternative view: an author is constituted by the reader as is a text, as much as an author makes the reader a reader.
Page 57
'5 7
I will highlight some of the relevant points from this perspective: Acknowledging that meaning is produced as to correspond to what an author meant, does not necessarily imply that it corresponds to what "the author" meant, not even that there is "the author". The reader says it because slhe believes an author would have also said so, that is, the reader has a justification for that statement which slhe believes would be acceptable to an author. Whatever the author says, slhe will do so believing that slhe has a justification for saying it which is acceptable to a reader. The author does not say whatever statements because a reader demarcates enunciation. An author does not produce meamng, but legitimacy, 1.e., a demarcation of meaning production. A reader does not produce indiscrimnate meanmg. Meaning IS produced, not truly correponding to what the author said, but by the very enunciation.
Therefore we can say that the two processes - meaning production and enunciation - are very close, in both cases demarcated by interlocutors: an author or a reader. The key point is that we produce meaning in order to belong to a social practice, to a culture, as much we produce enunciations for the same reason. While meaning is produced from a text, enunciation generates a text: setting the two processes apart. In my research I shall study a process of meaning production, by teachers, from ICT. Intellectual development is coming to be able to produce meanings which previously could only be produced with reference to someone else's authority: intellectual development is autonomy in the sense of an ability to anticipate acceptability, and that is achieved through the intemalisation of interlocutors.
Research Questions
(1) Can the study of meanings teachers produce from an educational software explain the way
they use it? (2) How can the study of designer (author), software (text) and teacher (reader) contribute to the use of ICT in schools and teacher education?
Page 58
Hypotheses What potencial the teacher sees in a software explains the way s/he uses it, and that is the basis for developing her/his own teaching approach. Analysing what the teacher sees in a software is a necessary condition to properly intervene on their teaching.
Comments This report is related to the short presentation of my early thinking on this research. I am currently designing a pilot study consisting of two mathematics teachers. I shall intervew and videotaping them working with Cabri. I have not described my research methods here as I am working on that and also I have not presented it in my talk.
Reference
Bottino, R. M. and Furinghetti, F.(1994): Teaching Mathematics and Using Computers: Links between teachers' beliefs in two different domains. J-F. M. J-P da Ponte (eds.). Proceedings of the PME XVIII, Lisboa, Portugal.
Cooney, Thomas J.(1994): Conceptualizing Teacher Education as field of inquiry: Theorical and Practical Implications. J-F.M. J-P. da Ponte (eds.), Proceedings of the PME XVIII, 2- 225. Grint, Keith and Woolgar, Steve (1997): The Machine at Work: Technology, Work and Organization. Polity Press.
Lins, R. (1992): A framework for understanding what algebraic thinking is. PhD Thesis; Nottingham University, UK.
Page 59
Page 60
PRIMARY CHILDREN'S IMAGERY IN ARITHMETIC Sandra
Pendlington School of Education, University of Exeter This paper discusses four ideas about imagery drawn from the domain of psychology and illustrates these ideas using examples from pilot study interviews with primary children.
Introduction
Gray and Tall describe strategies used by primary children to do simple addition and subtraction (Gray and Tall 1994). Low attainers were found to use predominantly counting strategies and, when compared with high attainers, are described as 'doing a different kind of mathematics that is intolerably hard', My own experience as a teacher of low attainers in years 5 and 6 agrees with this statement. I have observed the struggle these children have in maintaining a counting strategy. To help the children move to more flexible strategies, I devised a programme based on the manipulation of images of numbers. This programme had some success but the reasons for this were not clear.
My current research follows on from this and is centred on gaining an understanding of the nature and use of imagery by primary children and how use differs according to level of attainment. My research is in a very early stage. I will consider four ideas drawn from my literature survey of psychological research. Examples from early pilot work will be used to examine the connections between the psychological research and primary children's arithmetic.
Page 61
What is an image?
Kosslyn (1994) defines images as 'internal representations that "stand in
for" (re-present) the corresponding objects' and Horowitz (1978) as 'any thought representation that has a sensory quality'. The two points to
emerge from these definitions are that images are mental representations of actual perceptions and that they involve sensory experiences. 7. With these criteria in mind are the following examples of imagery? A drawing of dots to represent a million A description of subtraction using a decomposition algorithm A description of addition using a 'flying arrow' (fig. 1) Silent table chanting
Of the four, the drawing of dots seems to be the most likely image. However, it is not based on an actual perception because the child had never seen a million dots. The drawing is based on a mental picture of a million as a large number. The taught algorithm and flying numbers were descriptions of mental 'pictures' of how the tasks were carried out and so could be classed as images. Both were produced by high attainers and rely heavily on symbols. The interviews showed that high attainers were able to manipulate 'mental symbols'. Others required pencil and paper to produce a result, so moving the task from a mental to an actual manipulation. The final example is given as a possible use of an auditory image. This adult, when asked 7 x 8, backtracked to 3 x 8 and then silently chanted to 7 x 8. In these examples, images are used when a mental
manipulation takes place based on original visual or auditory sensory inputs. Imagery in action The idea of mental manipulation of images is supported by Kosslyn's (ibid.) theory of image generation (fig. 2).
Figure 2: Kosslyn's theory
One adult described 'scanning the sum to see whether it would be difficult '. William described getting a picture of a sum without noticing individual numbers, then slowly focusing on the numbers. In both cases the image was maintained during inspection. Transformation of an image can be seen in Lily's description of 8 + 5 (fig. 3).
Figure 3: Lily's 8 + 5
Lily needed to generate the dot patterns for 8 and 5, maintain them while
she considered what manipulation was needed and then transform the 8
into a 10. Transformation is a dynamic stage with the images flowing from one to the other rather than one disappearing and the next reappeanng.
Imagery as a knowledge-weighted process

Intons-Peterson and McDaniel (1991) discussed the role of knowledge in the use of imagery. They said that 'the more familiar we are with a task, the
more likely we are able to create an elaborate "knowing image '''. They
also stated the opposite 'the less familiar we are with a task, the closer the
imagery is to the original perceptions'. Looking at examples from the pilot

study work, adults showed more creativity than children. F or example when working with subtraction, the children tended to follow each stage exactly as taught, whereas adults jumped or combined stages. High attainers were more creative in their arithmetic methods, for example using 'difference methods' for subtraction (fig. 4).
Figure 4: Claire's subtraction 1050 - 670
In this example Claire, generates an image of a line, partitioned at 700. She then focuses first on 700 to 1050, then on 670 to 700. Finally she regenerates the whole image to add the two sections. Kosslyn's stages can be clearly seen as she maintains, inspects and transforms the image but the whole process is based on a creative use of her knowledge of the number system and her understanding of the subtraction operation.
The interviews yielded examples of children using their everyday knowledge to assist them in forming images. Images for minus 4 included
a thermometer, throwing snowballs, sliding on ice and goose bumps. The latter was accompanied by shuddering, the only example I found of a kinaesthetic image. Imagery as a mediator in learning Paivio (1971) describes imagery and language as 'mediators in learning'. Mediation could be thought of as scaffolding. Images are helpful in helping children to make sense of what they are learning and to remember what they need to know. Using the definitions given earlier, imagery is a mental activity and therefore would be a form of self scaffolding in the later stage of Vygotsky's Zone of Proximal Development (Tharp and Gallimore, 1988). Lily's dot images were being used as such a scaffold. These images had developed after specific image teaching (assisted learning in the earlier stage of the ZPD). Another example from the pilot study shows an image spontaneously created by a child and then used consistently in subsequent work. Daniel wrote 101 as 1001. During help to correct this problem, he suddenly pointed at the final zero of the hundred and said 'It's like a bomb exploding and becoming a stick!' The drawing in figure 5 followed. It was this visual image rather than the verbal description that was retained for further use.
Final thoughts Images are internal thought processes and as such are hidden from view. The external expression of images, whether a verbal description or put on paper, does not necessarily match the internal processes. The pilot study has shown that there are many problems in validating images, particularly those of very young children. This is why the images presented in this
bS
paper are those of older children (aged 9, 10 and 11) and adults. The technique associated with grounded theory has been used to check data, both by use of selected tasks and by direct checking of drawings and statements. At this moment in time, work with younger children is still being developed.
Even with this major qualification, I have found that the psychological research considered in this paper could be linked to examples in my pilot study. References Gray, E. and Tall, D. (1994) Duality, Ambiguity and Flexibility: A "Proceptual" View of Simple Arithmetic in Journal for Research In
Mathematics Education 25(2) 1994, National Council of Teachers Horowitz,

MJ. (1978 2nd edition) Image Formation and Cognition. New York: Springer-Verlag Intons-Peterson, MJ. & McDaniel, M.A. (1991) 'Symmetries and
Asymmetries between Imagery and Perception', in Comoldi, C. & McDaniel, M.A. (eds) Imagery and Cognition. New York: SpringerVerlag Kosslyn, S.M. (1994) Image and Brain. Cambridge Ma: The MIT Press Paivio, A. (1971) Imagery and Verbal Processes. New York: Holt, Rinehart
& Winston Inc.
Tharp, R. & Gallimore, R. (1988) 'A theory of teaching as assisted performance' in Rousing Minds to Life: Teaching, Learning and Schooling
in Social Context. New York: Cambridge University Press
Page 66
AN ACTIVITY APPROACH TO TEACHING AND ASSESSMENT. Chris Day, South Bank University
I have briefly indicated some key developments in the history of materialist dialectics and some principles of activity theory which follow from them. I have illustrated these general notions with some references to a program of teaching and dynamic assessment that I introduced in an earlier paper (Day
1998). I have presented video typescripts together with some quantitative and qualitative data from my
research to show that differences in ability between children from different socioeconomic areas were increased rather than decreased in the course of the teaching program. This was shown clearly in data from both static and dynamic assessments.
Some key developments in the history of materialist dialectics:
I1yenkov (1974) has traced some of the aspects of the history of materialist dialectics which have been important in the development of activity theory. Previous logic had proved to be incapable of tackling the problems posed by rapidly developing scientific theory. In particular, if it was to be of use in the
development of new knowledge, logic had to be able to tackle the problem of the identity of thought and
reality, or truth, in a changing world. Some questions posed for the logicians were clearly nonsensical. For example the question: "What is the distance between the syllable 'a' and a table" asks us to compare the two in terms of a third, space, but the syllable 'a' is not definable by its position in space. The problem of connecting theory and practice, however, was real and important. Descartes (1596 - 1650) had looked for such a third term under which we could compare knowledge with subject matter, ideas with things, so that the two could be theoretically related in a judgement. How, for example, could one compare 100 in the pocket with 100 in the mind, or a line on a graph with an algebraic formula? He concluded that there was no third term that could be found to relate them. Only "God" could explain the connection. Spinoza (1632 - 77) saw that the problem had been wrongly posed. There were not two objects, thought and substance, but only one. This was the real thinking person considered from two different opposing aspects. For Spinoza, the third term uniting concepts and reality was nature. If nature were only geometric forms,
then one essential aspect would be missing from nature. Thinking would be imperfect and mechanistic because thought would be missing from nature. Nature extends into space and thinks. A thoughtful woman
with 100 to spend could trace this material action mentally before carrying it out in practice. Planning and spending here are seen as two attributes of nature rather than as ideal and real properties of the woman.
10 Page 67 7
Kant (1724 - 1804) returned to the problem of developing logic. For him, also, logic had to be concerned with the problem of what thought is and how it is related to the world. If I asserted on the basis of past experience that all swans were white and then discovered a black swan, my whole idea of swan would have to be reassessed. A new broader definition would have to be produced to accommodate this discovery and this new definition would contradict the old one. All judgements of experience without exception have this synthetic character and have the right to subsume under a concept facts which contradict the original definition. An empirical generalisation is true "a posteriori" in that it can only say with certainty that 'all swans we have met so far are white'. Any inductive scientific generalisation will always be open to contradiction on the basis of new experience. General logic was thus competent only to judge analytical statements, in which the original idea was clarified and not synthetic statements which may be contradicted by experience. Kant proposed a "Transcendental Logic" of truth, in which synthetic judgements could be made and new predicates added to a concept without breaking the laws of identity and contradiction. I1yenkov (1974) has pointed out that if this experience is not considered in some arbitrarily determined part, but is taken as a whole, it will include a synthesis of judgements made according to categories that are not merely different but which are directly opposite and antinomic. The category of identity, for example, orients the intellect to invariance in the form of notions such as rate of flow of a river. The category of difference orients the intellect to the discovery of differences in objects seemingly identical. Infinite fluctuations could be discovered in the rate of flow of the river due to rocks in the river bed, the erosion and constantly changing contours of its banks and so on. One or other category must be ignored in making a judgement that is not internally contradictory and this problem recurs throughout all of Kant's categories. For Hegel ( 1770 - 1831), Kantian logic was still too limited. If it were to be a real science, logic would have to provide a much fuller concrete exposition of thought. For Hegel, thought had to be a rigorous reflection of activity and the sole task of logic was the clarification of the schemas of this reflection. Thought could not be presented to an observer immediately in the form of external or internal speech. It could not be identified simply with linguistic activity and logic could not be identified with the analysis of language. Man did not only function as a thinking being when talking. In fact, he often demonstrated his mode of thinking more adequately in his actions than in his explanations of them. For Hegel the forms of man's became embedded in materials in the process of man's activity and he consequently viewed history as a process of the embodiment oflogic. Marx (1818 - 83) substituted the notion of social activity for Spinoza's notion of "Nature". He argued with Hegel that the accumulated knowledge of generations should not be allowed to confront
Page 68
man as a collection of alien objects. This alienation could be overcome if the knowledge were appropriated by men as a tool for use in productive activity. Vygotsky (1934/1986) tried to limit the development of a science of materialist dialectics to the development of a synthesis of traditional and progressive notions of development. The notion of analytic induction forms the core of Piagef s theory of equilibration. A spontaneously occurring theoretical construct is tested in practice until a contradiction is found and the need to accommodate the contradiction leads to modification of the theory. A child will swat an irritating fly landing on her arm once and then will be able to do this for other flies. When a hornet arrives and she swats it she may well be stung and a new predicate will be added to irritating flying objects. Vygotsky has noted that a central problem with Piaget' s notion of equilibration is to be found in empirical studies showing that, contrary to Piaget' s assumption, awareness of difference or contradiction is actually easier to achieve than awareness of similarity because the latter awareness requires a more general notion under which the similar objects can be compared. It is easier to distinguish between a fly and a hornet than to form a general notion of a fly. A dialectical synthesis would have to incorporate both inductive and deductive processes as moments in the development of this theory. Gal'perin (1976b/89, 1980) pointed out that although this analysis was basically sound, unless activity theory could explain how activity was developed in practice it was in danger of returning logical development to rationalistic dualism. He subsequently investigated the logical development of thought as an integrated duality in the process of appropriation of orientation and executive actions. He argued that logical concepts (thoughts about thoughts) are developed from a combination of executive and checking actions to constitute what we later recognise as intuitive understanding and "attention".
In traditional teaching practices general mathematical notions such as "fractions" or "algebra" are often taught separately from practical applications with the result that students have problems developing a conceptual mathematical knowledge that can orient them towards the solution of problems they encounter. The search for the solution of a problem without this orienting basis for action can lead to an inefficient "trial and error" strategy in problem solution. Activity theory involves the use of the continually changing nature of the meaning of words or language to link together general and particular notions. In this process the student comes to see, remember, attend to and generally conceptualise the notions as they are encountered in practical problems in the same way as an experienced mathematician, such as the teacher, would conceptualise them. This includes
Page 69
appropriation of the rational form of the quantities. It must also include an orienting base for the actions to enable the student to use this ability in solving problems. An integrated program of teaching and dynamic assessment. I prepared a series of lessons based on a translation I made of a study by G. Nikola, presented in books by N Talyzina (1981,1995). This program modelled the notion of "Rate" as an introduction to the theory of mathematical functions. The rational form and a scheme for the orienting base were presented in written form in cards and in trees of reason. These two aspects of the teaching were introduced in written form so that they could be used immediately without artificial memorisation. Problems were presented in a carefully prepared program designed to lead the student through a process of abstraction and condensation in which the language used came to embody executive, orientation and control functions as conceptual models of the notions presented in the cards. Stage one involved the checking and consolidation of basic knowledge, the use of linear and graphical models and an introduction to the novel form of the lessons, including the use of card 1 to introduce work in forming an orienting base for actions. This was essentially a system of instructions which demonstrated how actions should be carried out but did not yet enable the pupils to act independently. Stage two involved the development of actions in materialised form with extensive specification of all orientation, executive and control functions for all variables. Cards one and two were used in the solution of examples that included all possible combinations of the three general notions of time, rate and product of the processes. This stage also included presentation of a wide variety of examples in a way that required some combination of both material and verbal activity. In particular this involved the use of examples with missing variables that were therefore impossible to solve, extra variables that had to be consciously ignored and the construction of examples by the children themselves.
In stage three actions were developed in the form of (written or spoken) external speech. These were
still neither automatic nor condensed actions. Card three was introduced together with action diagrams built into trees of reasoning in order to enable the formation of more complex actions, which could be pictured immediately through a number of progressively more highly focused layers of definition. Examples at this stage were designed to ensure that the speech became an independent embodiment of both general notions and concrete actions. For example, Vi would come to be seen automatically as the rate of output of a particular pump or the wage of a person, depending on the question in which it appeared. The children were required to continue the practice of constructing their own examples and abstract practice was introduced both as the object of certain lessons and in all subsequent ones.
1 0 From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 70
In this way they moved 011 to stage four of the process in which actions were developed as unvoiced external speech. Actions began rapidly to experience condensation and automation and acquired a form of operation conforming to an algebraic formula. It was possible for a student reaching this stage of development to reply immediately, for example, that a man running at 4 mps for 10 seconds would cover a distance of 40 metres. Stage five of the appropriation process, in which the actions became so automatic that they were inaccessible to introspection, was clearly reached in many cases but I stopped at this point to investigate proximal development zones. Qualitative and Quantitative Assessment. I have developed a two-stage dynamic assessment procedure that will partially quantify this learning process. The lessons are recorded on video in order to monitor conceptual development. Following teacher explanation, the researcher focuses on a particular student and helps with the work slhe is doing. The video record is reviewed immediately after the lesson so that adjustments to the individual work program can be made before the next session. In the dynamic assessment, following a course of instruction that provided a well defined context, clearly understood by both student and teacher, the researcher sits with each child for about an hour and works with him or her to complete a discourse structured around the completion of a theoretically based series of practice papers. In this way we can obtain both qualitative and quantitative measures of conceptual development (as opposed to notional understanding) and hence evaluate the success of the teaching process in developing problem solving ability within the field of instruction (Day 1998).
Qualitative video analysis showed a marked change in attitude from the initial introductory stages one and two, which were more traditional in form, to the second phase in which conceptualisation was developed. There was a great increase in interest and involvement in all three schools and a level of enjoyment and value attributed to the work that I have rarely seen in English schools. Children in the study were aged eleven or twelve and were considered by their schools to be working on levels three to five of the national curriculum. They scored on average 63% in a level 3/5 SAT paper at the beginning of the program. Problems set in topic tests t 1 and t2 at the end of the program were of the sort presented at GCSE levels 7 - 10. Scores ranged from 42% on average in test 1 to 52% on average in test 2, after one hour of individual tuition. I stopped the teaching process before the stage of 'fluency' was reached by most pupils in order to carry out the dynamic assessment. The children from the "poorer" schools (1 and 2) started behind their peers and fell further behind on both static and dynamic measures. The charts on the next page show static and (inverse) dynamic measures of average performance in three schools at the end of the teaching.
7\
Static test scores for schools 1 - 3
Dynamic scores (measuring inversely) for schools 1 - 3
Although the power of activity theory to generate innovation in teaching and assessing mathematics has been clearly demonstrated, there were a number a children who failed to achieve the target set for fluency in the domain. This failure was due partly to insufficient flexibility in the teaching program which could meet the needs arising from the social variation built into my sample. In particular the central aim of ensuring successful completion of checking and control aspects of the activity proved problematic because of this variation. It seems likely that a more sophisticated individualised teaching program using techniques from established schemes such as 'SMILE' mathematics will be needed to overcome these problems and research to develop such a program is currently under consideration. References DAY, C. (1998) Making Sense of Modelling with Mathematical Functions: A Rational, Reasonable and Rigorous Approach (British Society for Research and Learning in Mathematics). (In press). GAL'PERIN, P, Y A. (1980). The Role of Orientation in Thought (In Soviet Psychology Vol 18 part 2 pp 84-99). GAL'PERIN, P. Y A. (1976b/89) The Problem of Attention in Thought (In Soviet Psychology Vol 27 part 3 pp 83- 94). IL YENKOV, E. V. (1972) Dialectical logic: Essays on its History and Theory (Moscow: Progress Publishers) TALIZINA, N.F. (1981) The Psychology of Learning: Theories of Learning and Programmed Instruction (Moscow: Progress Publishers) TAJIbI3.HHA, H.$. (1995) $OpMHpoBaHHe [lpHeMoB MaTeMaTH4ecKoro
MhllliJleHrU{ (MocKoBcKHH rocy~apCTBeHHblH YHl1BepCHTeT HM M. B.JIoMoHocoBa) VYGOTSKY, L.S. (I986) Thought and Language ( Translated by Kozulin,A: Cambridge, Massachusetts MIT Press).
( L From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 72
Mathematical Support for Engineering and Science Students

Peter Gill King's College London
An investigation of the mathematical difficulties experienced by undergraduates starting courses in science and engineering has revealed a range of misconceptions that appears not to have changed greatly over several decades despite changes in school curriculums. Many of these difficulties relate to the understanding of graphs. There is also evidence that the students find the style of learning expected at HE to be greatly different fromwhat they have experienced at school and this too contributes to their problems. A number of reasons for the problems are hypothesised and some solutions are proposed.
Introduction
In the wake of the publication of the Engineering Council report on the changing mathematical background of engineering undergraduates (Sutherland & Pozzi, 1995) we have been investigating the issue within King's. This has involved one education lecturer working with first year undergraduates of physical science and engineering who have been identified as being likely to have difficulties with their mathematics. The initial identification is made using a diagnostic test developed some time ago for such work and modified over several years. The selected students then meet in groups of about a dozen for an extra maths tutorial each week. The content and style of the tutorials are informed by their performance on the diagnostic test and draw on recent developments in mathematics education research and ongoing research taking place in the School of Education, especially CAME (Cognitive Acceleration in Mathematics Education)
Background
Worries about engineering and science students' mathematical competencies are nothing new. Over a quarter of a century ago lecturers in the King's maths department wrote "Many students of science subjects arrive at university with little facility and less interest in mathematics" when describing the reason for setting up a system of mathematics support tutorials (Baker, Crampin, & Nuttall, 1973), and even Sir Ron Dearing in his Review of Qualifications for 16-19 Year aIds commented" ... representations by employers about standards in ... the application of number have been a feature of national life for more than a century. It has always seemed that things were better twenty years ago" (Dearing, 1996)p6. Descriptions of the areas of mathematical weaknesses over the years have had much in common: calculus, especially integration, always appears. Other topics lists regularly include logarithms, trigonometry, complex numbers and algebraic manipulation especially factorisation.
The stability of the criticisms over many decades accompanied by a clear failure to address them, led us to believe that many of the problems were at a much more basic level. Even this is not a new hypothesis, in a paper on the mathematical deficiencies of engineering undergraduates written nearly twenty years ago the authors, having produced a list such as that in the previous paragraph, write
Page 73
"Our evidence is that the deficiencies are more fundamental than this list would suggest or than students are prepared to admit" (Howarth & Smith, 1980). Accordingly the diagnostic pre-test addressed mostly topics that traditionally are taught at pre-sixteen level and contained a number of tasks designed for national surveys of school mathematics (e.g. CSMS and APU). The topics covered included simple algebra and number, the laws of indices especially numbers in standard form, and quite a few questions on the understanding of graphs both visually and in terms of y = rnx + c.
Findings Despite the students' having passed mathematics examinations at 18+, there was evidence of a range of
fundamental misunderstandings:
(i) A number of the diagnostic test questions were designed to elicit standard misconceptions as displayed by pre-sixteen pupils. Many of the students still displayed exactly such misconceptions. (ii) Students could often show success in one area but fail when the context or style of the problem was changed. There is evidence of the learning of mathematics as a set of disconnected skills in order to pass examinations. (iii) There was surprisingly little difference in overall pre-test performance between engineering students with BTEC, physics students with A-level maths, and chemistry students with no more than GCSE maths. However it seemed that the post A-level students were generally slightly better at graph interpretation than the post BTEC students. We suggest that this is because most A-level maths syllabuses require the use of graphical calculators which is not usually the case for BTEC. On the other hand the post BTEC students seem better at number problems. We suggest that this is because they are generally older and have more experience of using number. (iv) The most significant area of skills and knowledge in terms of predicting success in the university mathematics examinations for the engineers concerned graphs, including the interpretation of shape (especially changes of gradient), verbal interpretation of graphs and the non-calculus algebra of graphs. This connection was extremely strong and further investigation during tutorials and interviews has supported its importance. In many ways this is a surprising conclusion because the end-ofsemester examination concerned had apparently virtually no requirement for these skills. We hypothesise that we may have uncovered an underlying cognitive structure that is necessary for success at higher level mathematics. This is being investigated.
Page 74
(v) Many students are unsure of how to learn mathematics and find the jump from schoollFE to HE to be very great. In a large university the opportunity for individual or small group tuition is limited. King's uses the standard style of large formal lectures backed up by smaller examples classes. However these smaller classes may have up to 30 students in them. For many students, learning is seen as a passive activity and this style of teaching tends to reinforce that view. When asked (in the support tutorials) to talk about what they are doing or to work in pairs and small groups they find it very uncomfortable. There is little evidence of ability or willingness to reflect on their own learning so in the wake of the success of such techniques in the CAME (cognitive acceleration in mathematics education) project, much time is spent helping the students to develop such metacognitive skills.
Discussion
Perhaps the most remarkable finding is the stability of the problem areas despite massive changes in the mathematics curriculum at school level. Twenty years ago school mathematics syllabuses were designed as feeders for university mathematics. With the introduction of GCSE, the National Curriculum and modular A-levels this is clearly no longer true. The mathematics curriculum up to GCSE is very much broader than in the past and is designed generally for purposes other than preparation for HE maths. Such changes, accompanied by the much wider recruitment base of undergraduates should lead to the expectation of a very different range of mathematical problems to that encountered twenty and more years ago. The fact that this seems not to be the case is of considerable interest.
This is not to say that there haven't been any changes. Comparison between the test scores of the students entering in 1995 and those of students sitting the test in 1983 show that there have indeed been a number of changes. However subsequent observation and discussion with students tends to indicate that the changes are in the areas of skills rather than underlying concepts. In other words the widening of the school maths curriculum has given students less opportunity to practice mathematical skills (or rather, those mathematical skills required for the HE maths courses under discussion.) because of the range of mathematical study required. For instance the National Curriculum for mathematics includes a considerable amount of statistics of the kind that was only met on undergraduate courses twenty years ago. At the same time this increase in content may have left less time for teaching for conceptual understanding.
Another factor is that of teaching to the test. Teachers have always done this to some extent and it is unavoidable. However the introducing of criteria for GCSE and the National Curriculum accompanied by a tightening up of examination rules makes this easier. In addition because of the introduction of teacher assessment and coursework, many teachers are more aware of assessment theory than in the past. This is producing cohorts of students who are good at passing certain styles of examination regardless of content. See (Gill, 1994) for evidence of this.
Page 75
7 G
There is also the issue of what service mathematics is for. It is quite clear that the students see no relationship between their mathematics courses and their main subject studies. An example of this was revealed in a discussion with a group of physicists at the end of their first term at university who were asked how their studies in mathematics related to their studies in physics both at A-level and at university. They were unanimous that there was no apparent relationship between the two subjects at either level. While they agreed that the maths course they study at university would be impossible without A-level maths, they made no other connection. Indeed the most remarkable quote was " ... we haven't used maths in Physics much", a comment which raises questions about what they think mathematics actually is. Further probing revealed that they didn't see the quantitative and algebraic content of physics as being mathematics.
This situation is exacerbated by the messages, implicit and explicit, that the students receive from their lecturers. It is not an exaggeration to say that they are caught in the crossfire of a three way battle for the mathematics curriculum both within the institution and on the wider scale. The views of mathematics held by the mathematics establishment, the mathematics education establishment, and the users (science and engineering departments), are frequently mutually exclusive and many of the lecturers concerned are not reticent about expressing these views in public and to their students.
Where now?
The first thing to realise is that rapid wholesale changes in the maths curriculum at school level are simply not going to happen for political reasons, and anyway this investigation is revealing that many of the problems are curriculum proof. However there are some promising avenues. The cognitive acceleration schemes, CAME and CASE alluded to earlier (Adey & Shayer, 1993) have had considerable success in raising cognitive levels and examination success in GCSE mathematics and science. Indeed industry (BP) has put funds into the area by sponsoring the training materials for CASE. One of the reasons for the success of these schemes is that they require considerable professional development in the teachers concerned over an extended period (at least two years). Unfortunately that is very expensive.
To reduce the amount of teaching to the test a wider range of assessment techniques need to be introduced at all levels. The frighteningly poor reliability (using the term in its technical sense) of standard written examinations needs to be more widely appreciated. It is perhaps significant that the previous government's assessment agency refused to carry out reliability studies on its National Curriculum tests. One way forward would be to raise the status of Teacher Assessment of the National Curriculum. However this too would require more professional development for teachers with its accompanying funding implications. The scheme of verifiers used in GNVQ would make a good starting model for this.
1 6
Page 76
There have always been problems of discontinuity between different phases of education. The introduction of the National Curriculum improved the primary/secondary link, however there continue to be claims that the shift from GCSE to A-level is too great. In fact research carried out for SCAA (Brown, 1995) indicates that this is not such a great problem as had been thought. There has been little or no recent research into the shift from school to vocational courses, however some work done twenty years ago with vocational students (Furneaux & Rees, 1978) indicates similar problems to those found in the present investigation. Where there clearly is a problem is between collegeNIth form and HE. Many school teachers are unaware of what happens at university and there is no doubt that there is considerable lack of knowledge among HE lecturers about what is happening at lower levels. This may partly be due to lack of interest and head-in-the-sandism, but is certainly exacerbated by the scale and rapidity of the changes at school level at a time when HE itself is in upheaval. Some recently published research into the problem of mathematics for engineering students (Crowther, Thompson, & Cullingford, 1997) identifies this lack of continuity and information asthe major factor at play.
The basic content of HE ancillary maths courses has not changed in a generation but incoming students' knowledge has. It has been suggested (Ireson, 1996) that the situation could be improved by considering what the students entering HE comewith from their earlier studies. The example of their knowledge of statistics has already been quoted but it is little realised just how IT literate our current students are. Students now entering HE are among the first to have followed the National Curriculum for IT and in many cases they are well ahead of their lecturers. Spreadsheet mathematics has entered most fields and all the new students have experience of it. They find it very odd to be told that the only IT used in the HE maths courses is the standard scientific calculator, not even graphical calculators being permitted in many university departments.
Finally The stability of incoming undergraduates' misunderstandings over many decades of change has a number of
implications, not least that the evidence of educational research is not being taken on board by teachers and textbook writers and perhaps we in the BSRLM should spend more time considering how to extend our influence. However at this stage the biggest problem revealed by this investigation seems to be the lack of cross phase dialogue between school and HE and perhaps this is where to start.
Page 77
I (
References
Adey, P. & Shayer, M. (1993). Really Raising Standards: cognitive intervention and academic
achievement. London: Routledge.

Baker, J.; Crampin, M. & Nuttall, 1. (1973). A crash course in calculus. International Journal of
Mathematics Education in Science and Technology, 4,335-339.

Brown, M. (1995). The Step between GCSE and A-level in Mathematics. Report prepared for SCAA. Crowther, K.; Thompson, D. & Cullingford, C. (1997). Engineering degree students are deficient in mathematical expertise - why. International Journal of Mathematics Education in Science and
Technology, 28(6), 785-792.

Dearing, R. (1996). A Review of Qualifications for 16-19 Year oids: Full Report. Report prepared for SCAA. Furneaux, W. D. & Rees, R. (1978). The structure of mathematical ability. British Journal of
Psychology, 69, 507-512.

Gill, P. N. G. (1994). Comparisons of Key Stage 3 pupils' performances in national tests in Science and Mathematics: a question of validity. British Journal of Curriculum and Assessment, 4(2). Howarth, M. J. & Smith, B. J. (1980). Attempts to identify and remedy the mathematical deficiencies of engineering undergraduate entrants at Plymouth Polytechnic. International Journal of Mathematics
Education in Science and Technology, 11(3),377-383.

Ireson, G. (1996). Improving the links between the teaching of physics and mathematics in the 16-19 age
range. Report prepared for Institute of Physics (unpublished).

Sutherland, R. & Pozzi, S. (1995). The changing mathematical background of undergraduate
engineers: A review of the issues. Report prepared for The Engineering Council.
7 Z From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 78
TUTORS' REFLECTIONS UPON THE DIFFICULTIES OF LEARNING AND TEACHING MATHEMATICS AT UNIVERSITY LEVEL: A REPORT OF WORK-IN-PROGRESS
Elena Nardi University of Oxford

ABSTRACT
Following a doctoral study on the learning difficulties of20 Oxford first-year mathematics undergraduates in their encounter with mathematical abstraction, a study is currently being carried out in which the mathematics tutors who participated in the initial study have been asked to reflect upon samples of its data and findings in semi-structured interviews. The interviews address three areas: clarifications and explanations regarding the pedagogical and psychological language used in the samples, a validation/critique of the interpretations in the thesis regarding the students' learning difficulties and a reflection upon the events in a tutorial from a teaching point of view. The focus of the discussion alternates between a specific (sample-centred) and a general addressing of issues relating to the learning and teaching of undergraduate mathematics. Here short extracts from the interviews are presented and commented upon. Transcription and analysis are now in progress. The research reported in this paper is a brief follow-up * to the author's doctorate (Nardi 1996) and a pilot to a one-year ESRC-funded project due to begin in October '98 (Jaworski and Nardi 1998). In the following I present the doctoral thesis briefly; then the aims, methodology and data extracts of the current project; finally I link this with the upcoming research. 1. BACKGROUND: THE THESIS
The doctorate (Nardi 1996) set out to explore and elaborate upon the documented view that abstraction is possibly the main inherent epistemological characteristic of mathematics to which its didactical complexity can be attributed. The study originates in the idea that, in order to formulate a didactical discourse on advanced mathematics, microscopic psychological studies are needed which identify the students' needs and then engage in trying to cater for them. The doctorate was a qualitative study of 20 first-year mathematics undergraduates' learning difficulties in five mathematical topics: Foundational Analysis, Calculus, Linear Algebra, Topology and Group Theory. The students were observed for two terms during the tutorials given to them once a week on a 1: 1 or 2: 1 basis and were interviewed at the end of each term. The minimally participant observation sessions and the semi-structured interviews were audio-recorded. The recorded material was analysed, using a combination of techniques from Data Grounded Theory (Glaser and Strauss 1967) and Discourse Analysis (Dijk 1985) in terms of the students' explicit or implicit articulations of their difficulties. In the analysis, dominant theories in the field of the Psychology of Advanced Mathematical Thinking (an
* funded by the Harold Hyam Wingate Foundation
Page 79
account of most of these theories can be found in (Tall 1991 and a theoretical framework that coordinated aspects of the developmental theories of J. Piaget - specifically his concept of Reflective Abstraction (Dubinsky and Lewin 1986) - and the sociocultural theories on learning of L. Vygotsky (1978) were employed. The aim of the study was to offer a psychological profile of the students' difficulties across the above mathematical topics. The novices' encounter with mathematical abstraction was described as a personal meaningconstruction process and as an enculturation process: the new culture is Advanced Mathematics introduced by an expert, the tutor. The novices' concept-image construction was described as a construction of meaningful metaphors and an exploration of the 'raison-d'-etre' of the new concepts and the new reasoning and was characterised by the tension between the informal/intuitive/verbal and the formal/abstract/symbolic - which was discussed in terms of semantics and reasoning. The novices were in difficulty with the mechanics of formal mathematical reasoning as well as with applying these mechanics in a contextualised manner. This decontextualised behaviour was linked to the fragility of their knowledge with regard to the nature of rigour in formal mathematics. Issues related to teaching were touched upon when it was deemed necessary to the analysis of the learning instances that had been selected.
Some of these issues inform the theoretical basis for the current research. On the basis of the evidence in this study, to integrate teaching styles that the students are familiar with from school into their initial experiences at university level and to present mathematics in a way that is more revealing about the mechanisms that characterise mathematical thinking seem to carry significant didactical potential. From this study several conjectures relating to teaching mathematics at university level emerged and can be seen as indicators of good practice: A to clarify the rules of the formal game of advanced mathematics, that is to provide clarity with regard to the increased requirements of rigour in the new course that the novices have to confront - for instance with regard to the knowledge that they are allowed to assume; B to co-ordinate intuitive and formal practices, yield didactical control to the learners and view interaction with the students at a reflective, meta-topical level as a legitimate part of the teaching; C to explore the varying efficiency of types of visualisation and to develop flexible metaphors to support the understanding of abstract concepts; and, D to support the learners' existential, meta-mathematical enquiries and to launch new concepts as epistemologically useful and necessary apparatus. In the research, currently being conducted, the tutors are being invited to reflect and comment upon samples of data and analysis relating to the findings of the above. In the following section I describe
~o
briefly this research as a bridge between the strictly psychological concerns of the doctorate and the directly pedagogical concerns of the upcoming ESRC-funded research. 2. THE CURRENT PROJECT
The research described in this section is a short follow-up of the doctorate described above and is currently being conducted. Its primary aims are:
to provide feedback to the tutors who participated in the doctorate, to enrich the findings from the doctorate, by including the participant tutors' point of view, to introduce a pedagogical dimension in the psychological discourse developed in the doctorate, and to inaugurate the collaboration between mathematicians and mathematics educators involved in the proposed research. In this sense the project described in this section is a Pilot to the upcoming research, in the development of discourse and methodology.
For the above purposes, the plan, currently being implemented, is as follows:
Re-establish contact with the tutors who participated in the doctorate and request their collaboration. The request has been met with positive responses.
Prepare the feedback material to be presented to the participants: this consists of samples of the data,
transcribed extracts from the tutorials, and the analysis, presented in the doctorate, deliberately chosen so that they trigger off the tutors' reflection upon the students' learning processes as well as their response to the analysis.
Introduce the above aims to the participants and conduct semi-structured interviews in which they are
asked to react to a sample of data and analysis with regard to - their response to the researcher's interpretation of the students' learning processes in the sample, - their own interpretation of the students' learning processes in the sample, and - their reflection upon the teaching actions taken in sample.
Analyse the interviews. This aims to be the development of a discourse in which the analysis in the
doctorate and the tutors' interpretations, as expressed in the interviews, are juxtaposed. The rationale for this juxtaposition is that it allows the identification of the possible differences of perspectives between mathematicians and mathematics educators. Given that the upcoming research involves the collaboration between mathematics educators and mathematicians who teach at university level, a collaboration which, as explained in the concluding section, will take the form of a clinical
~\ Page 81
partnership (Wagner 1997), the doctorate, and in particular the current research, set up a forum for this collaboration: they allow establishing and modifying the methodological principles and clarifying the issues to be raised by the upcoming research. The interviews, which were audio recorded, are now being transcribed. I exemplify the tutors' reactions by citing extracts ** from the interviews and in particular from the parts of the discussion regarding area C of the implications for teaching as discussed in the doctorate and mentioned in Section 1, namely the varying efficiency of types of visualisation and metaphors as supports to the students' understanding of abstract concepts. Preliminary analysis suggests that the tutors' reactions vary from an educated confidence in their longestablished practices,
82.
An analytical framework which integrates the learning-oriented perspective of the doctorate with perspectives on the tutors' self-reflective processes is currently under development.
Page 83
3. THE FUTURE: THE ESRC PROJECT
The broad aim of the upcoming, more action-oriented I-year research is to explore, in a partnership with university mathematics teachers, current thinking and practices in mathematics teaching at first-year undergraduate level. Participants of the research will be Oxford mathematics tutors. The researchers will observe tutorials in which the participants are involved and then interview the participants. The analysis will produce documentation and grouping of practices, processes and the thinking of participants. Researcher interpretations will be fed back to participants for respondent validation but also, in the spirit of the intended partnership, to initiate discussion on the implications of the research findings for future practices. To exemplify how the tutors have expressed their expectations from the upcoming partnership, I conclude with two remarks made by the tutors in the current interviews:
REFERENCES Dijk, T. A. v. 1985. Handbook of Discourse Analysis Vols 1-4, London: Academic Press Dubinsky, E. & Lewin, P. 1986. 'Reflective Abstraction and Mathematics Education: The genetic decomposition of induction and compactness.', The Journal of Mathematical Behaviour, 5, 55-92 Glaser, B. G. & Strauss, A. L. 1967. The Discovery of Grounded Theory: Strategies for Qualitative Research, New York: Aldine de Gruyter Jaworski B. & Nardi E. 1998. Characterisations and Issues of Mathematics Teaching at Undergraduate Tutorials. Research proposal accepted by the Economic and Social Research Council. Award Number: ROOO 22 2688. Nardi, E. 1996. The Novice Mathematician's Encounter With Mathematical Abstraction: Tensions in Concept-Image Construction and Formalisation. Unpublished doctoral thesis. University of Oxford. Tall, D. 1991. Advanced Mathematical Thinking, Dordrecht / Boston / London: Kluwer Academic Publishers Vygotsky, L. S. 1978. Mind in Society: the Development of Higher Psychological Processes, Cambridge, MA: Harvard University Press Wagner, J. 1997. 'The Unavoidable Intervention of Educational Research: A Framework of Reconsidering ResearcherPractitioner Cooperation.', Educational Researcher, 26 (7). 13-22
Page 84
CONVENTION OR REALITY?
M. M. RODD & M. R BARBER The OPEN UNIVERSITY & THOMAS TELFORD SCHOOL * *from Autumn 1998: Leeds University & Cockshut Hill School, Birmingham, respectively. Abstract: this report both introduces to a mathematics education audience some research in contemporary philosophy of mathematics and develops the relevance of such research to mathematics in education. The philosophical theories to be considered are those of the physicalistrealists, specifically in this paper, Resnik and Bigelow; physicalist-realists offer theories of how mathematics is, in some sense, part of the physical world The notion of an 'instantiation' of a mathematical theorem, or object, is distinguished from a more general and familiar 'representation' in order to help conceptualise the possibility of non-linguistic experience of mathematical relationships or entities. The question of relevance to mathematics in education is approached by presenting mathematical activities to engage with and school-texts to analyse. Introduction Our objective, in this session, was to 'report' on work on realism in the philosophy of maths through workshop-style activities and discussion. The wider aim was to come up with an appropriate sense of the distinction between 'real' and 'conventional' by working with an example from the secondary school curriculum. The motivation for the research came from our sense, as mathematics teachers, that there is a difference between mathematical conventions (data representations, particular measures, rules for rounding up, for example), which are properly part of the school mathematics curriculum, and aspects of mathematics, also within the curriculum, the truth of which is independent of mathematical conventions people have established through the ages (for example, the existence of non-ratio-like numbers which we consider, below). But what is a suitable conceptual distinction between convention and reality as they pertain to the school mathematics curriculum?
To answer this question, we need to conceptualise 'convention' and 'reality' in broader terms, then see where parts of the curriculum sit (we have already exemplified this distinction just in order to clarify the basic question). The 'conventional' category is essentially the linguistic-cultural aspects of mathematics. And is currently the dominant conceptualisation of the nature of mathematics in education. In order to follow though on our practitioner's sense that there was something other than humanly constructed conventions, we looked for theoretical underpinnings for the notion of independent mathematical reality. And then we interpret this theory for the mathematics of the school curriculum.
Page 85
Philosophers who hold that there are convention-independent mathematical truths are called 'realists' in the philosophical register. And their conception of the nature of mathematics offers different insights from their contraries, termed 'anti-realists' by Michael Dummett. Anti-realists also include 'conceptualist' philosophers, like Rorty (1990) and Bloor (1976). The views, at least of the latter, have been interpreted in the mathematics education domain by Ernest (1991, pp 45-58). The theoretical underpinning we put forward was based on 'physicalist-realism'!. Physicalist-realists claim that mathematical entities do exist and their reality is bound to human beings' particular physical make up within the physical world. Although a sense of the 'objectivity' of mathematical entities can be explained without advocating this sort of reality (see, for example, Ernest, 1997), we choose to grasp the nettle of the 'objectivity' of mathematics directly by this realist ontology. The specific philosophers we quote here are Michael Resnik (1993, 1998) and John Bigelow (1988). A brief introduction to realism in mathematics Realism in mathematics, within philosophy, concerns ideas about how mathematical 'abstractions' may indeed exist. There are many forms of realism, from Plato's conception of transcendent 'Forms' to Putnam's 'internal realism'z (e.g. Putnam, 1990). However, a collective feature of realist theories of the world is that "existence is prior to theory" (Harre, 1986, p5). In terms of contemporaries, Harre's 'modest realism' recognises that securing a scientific belief is, in some sense, a social activity. Nevertheless, Harre insists that "for there to be public reliability something must exist independently of whomsoever first found it." (pI2). Putnam eschews conceptual relativism like the Rortarian 'conceptualist' position3. Popper, too, was also a realist, (e.g.,1972). His metaphor of the 'mountain beneath the clouds' suggests a 'reality' which 'we seek'. In a nutshell: to assert 'realism' involves asserting the existence of an external world and it is that external world which, in theory, is the ultimate arbiter of truth values. Physicalist-realists in mathematics furthermore assert that mathematical existence presupposes a material reality which is the universe of which we are a part. In mathematics educationalliterature4, however, such positions have been over-shadowed in recent times by versions of 'conceptualism', which, roughly, sees mathematics as a language or social practice, (see Sierpinska and Lerman, 1996, for a review).
Irvine (1990) defines this type of realism thoroughly. position, briefly, denies the 'God's eye view' of nature: Putnam claims to be a 'small r' realist (Putnam, 1990). 3 "Rorty' s view is just solipsism with a 'we' instead of an I" (Putnam, 1990, pix) 4 At the June 1997 BSRLM, one of us, MMR, gave an introduction to Penelope Maddy's realist views, (see Rodd, 1998)
I
2 This
Page 86
CONVENTION OR REALIT'l1
Resnik and Bigelow Michael Resnik's view is that mathematics is a general science of patterns and structures (i.e. discovered results about patterns and structures). So, for example, whether a number is prime or not is a physical property of arrays of units (whether they can be made rectangular): "mathematical objects are abstract entities existing independently of us and our constructions and theories ... (most of) the claims of contemporary mathematics are true, and they are true independently of our holding them to be true ... mathematical reality transcends our own existence, beliefs and experience."(Resnik, 1998, p319) "We can never study a mathematical object, such as pi, in isolation, as we might study a specimen. Instead we study the patterns or structures in which mathematical objects are positions"(ibid., p326) Resnik denies that any part of mathematics is directly perceptible, (I 993, p50), but that mathematical entities are posited by limiting actions. His paradigm example being that of a geometric point, which is a posit derived from repeated cutting on a line. John Bigelow's view is that mathematics is the theory of universals, i.e. a systematic collection of statements about relations among universals5 "Natural numbers are, I urge, universals which are instantiated directly by individuals in the world; but the other kinds of numbers cannot be construed as properties or relations of individuals, but rather, must be construed as relations between relations, using a technique pioneered by Frege and Whitehead and Russell" (Bigelow, 1988, p5, emphasis in original) "sets as universals of a special sort, ... [for] all the structures mathematicians have hitherto wanted to study are instantiated by sets" (ibid., p6) "And since we are physical beings ourselves, through and through, the same mathematical patterns that are in the world around us are also present inside us" (ibid., p2) Bigelow's paradigm mathematics is that of the irrationality of .../2. We exploit his specific instantiation of .../2 to show the a-conventional existence of non-ratio-like numbers, below. Thus these two philosophers offer ways of locating some mathematics within the reality of the physical world.
Instantiation Resnik, as well as Bigelow, uses the term 'instantiation':
5 "universals are the properties relations, patterns, slructures and so forth, which can be shared in common with many diverse individual particular things" (ibid., p 11), for example, 'redness', '17ness', 'roundness', 'non-ratio-ness'
Page 87
"templates as concrete representations of patterns ... templates need not instantiate the patterns they represent; written musical score, for example, are templates for sound patterns"(Resnik, 1998, p326) Understanding the term 'instantiation' should help in the conceptualisation of a distinction between convention and reality in mathematics: an 'instantiation' is a special sort of representation which 'is an instance of the concept in question. Bigelow, above, was quoted as saying that numbers are the universals which are instantiated by individuals. A way of grasping the idea of physicalist-realism is to look for instantiations of mathematical theorems and to distinguish these from other sorts of representations (which are often cultural constructions). Instantiations are a sort of representation but a distinguished class which DIRECTLY represents the entity, relation or theorem. We offered an example of an instantiation for participants: Activity: to find an instantiation ofa square root algorithm (Devi, 1977, p71 ,Wheeler, 1974 p79) We start with a HTU square number: What is the square root of 289? 289 = 2xO + 8xl + 9 (where '0' stands for a one hundred square, and 'I' stands for a ten strip) The idea is to arrange the Os and the I s and the units in a square. Then the root is just the length of the side. How can this be expressed algorithmically? Then extended for larger numbers and for approximations to roots of non-square numbers?
Given this conception of instantiation, the next activity was to 'instantiatiate' the non-ratio nature of -./2. If this can be achieved, we claim that this shows the 'physicalisticaUy real' existence of an irrational number - a discovery, in short, rather than (just) a linguistic construction. The method for such an instantiation is worked through in Bigelow 1988, p33. The idea is to show that no square grid of pebbles - discrete, easily manipulable objects of perception - can ever be rearranged to make two
EQUAL
smaller squares. Thus the non-existence of two natural numbers such that n2 = 2m2 forces the
real existence of a non-ratio-like (a.k.a. 'irrational') number. Application to school mathematics Participants were given a selection of extracts from school maths texts (Bostock et al. 1993, Hackney et al. 1994) to scrutinise with the objective of observing and c1assitying where the texts presented irrational numbers as a 'real phenomenon' and where they presented them as an invented construction.
Page 88
CONVENT[ON OR RF-AUTY'1
As an example of these extracts, here is a part of Hackney et al. : "We know that ..J2 is about 1.414. But the Greeks had no calculators, and no decimal numbers. The only way they could talk about numbers like this was in terms of fractions that is, as ratios of whole numbers. They could say that one and a half was i. But they could not find any ratio of whole numbers that was exactly equal to ..J2, which was not surprising, because there is no such ratio .. ,. Some people say that Pythagoras came to a sticky end because he invented irrational numbers." (p91) Tn usual practice currently, the topic of irrational numbers is one in which the learner completes the GCSE-Ievel study without knowing what an irrational number is. So, as an application of being able to distinguish between 'convention' and 'reality', we suggest that it may be helpful to the learner to have the teacher clarifY the conventions-reality status of a new topic, even if there'll be a lot more to the topic in future study. Conclusion One of the fascinating thing about mathematics - from our perspective - is the way conventions do combine and extend mathematical reality: abstraction from a perceptual or experienced phenomenon (e. g. perception of discreteness or relations of kinship, respectively) has both scientific and systemic features.) Furthermore, on this science/language distinction: one is not always able to decide easily, for a given mathematical topic, whether it is an encoded aspect of the world (which was discovered) or a projected convention which was (implicitly) agreed within some group of language users. Abstracting (formulating key features), axiomatising (defining conventions), discovering (finding out: either about things or consequences of the defined conventions) and modelling (applications of any of these) are, in our view, all important aspects of the mathematical enterprise. This enterprise is rooted in the physical experienced world and grows from there in receptive and nourishing cultures.
To sum up: the relevance of realist philosophy of mathematics to school mathematics has been exemplified and developed through the case of irrational numbers: 1. we claim that irrational numbers exist as real entities which can be instantiated physically though not directly observed (as no measured length 'measured up' as an irrational) 2. the encounter students have with irrational numbers either tends to represent them as lengths which can never be measured completely accurately or as algebraic symbols, manipulated algebraically. 3. so the question, which is a philosophical one, of the essence of an irrational number, is relevant to education. This is because, in the school texts we have found in our research, the concept of an
..J2, for example, which are
Page 89
CONVF:NTIONOR REALm~)
irrational number is dealt with 'conventionally' (i.e. use so many decimal places - 1.414 - or use this symbol
-.J2 for the length of the diagonal of the unit square) rather than as a 'real' entity on which the
conventions (of truncation or symbolic representation) have been laid to help with communication of and work with of these numbers in applications.
references
BIGELOW, 1. (1988) The Reali~y o.lNumbers (Clarendon Press, Oxford) BLOOR, D. (1976) Knowledge and Social Imagery (Routledge and Kegan Paul, London) BOSTOCK, L. et al. (1993) STP Mathematics 5A (Stanley Thorns, Cheltenllam) DEVl, S. (1977) Figuring (penguin, London) ERNEST, P. (1991) The Philosophy o/Mathematics Education, (Falmer Press, London) ERNEST, P.(1997) 'Texts and the Objects of Mathelllatics' Philosophy of Mathematics Education Journal no. 10 (electronic version) HACKNEY, 1. et at. (1994) Heinemann Mathematics Upper Course (Heinemann Educational, Oxford) HARRE, R (1986) Varieties of Realism (Blackwell, Oxford) IRVINE, A. D. (1990) (Ed) Physicalism in Mathematics (Kluwer, Dordrecht) POPPER, K. (1972) Objective Knowledge (Clarendon Press, Oxford) PUTNAM, H. (1990) Realism with a Human Face (Harvard University Press, Cambridge MA.) RESNIK, M. D. (1993) 'A Naturalised Epistemology for a Platonist Mathematical Ontology' in RESTIVO, S. et al. (Eds.) Afath worlds: philosophical and social studies of mathematics and mathematics education, (SUNY Press, New York) (pp. 39-60) RESNIK, M. D. (1998) 'Proof as a Source of Truth' in TYMOCZKO,T (Ed.) (new edition) New Directions in the Philosophy of Mathematics (Birkhauser, Boston) RODD, M. M. (1998) 'Maddy's Mathematical Reality' For the Learning of Mathematics, 18,2 pp.34 - 41 RORTY, R (1990) Objectivity, Relativism and Truth (CUP, Cambridge) SIERPINSKA, A. and LERMAN, S. (1996) 'Epistemologies of Mathelllatics and Mathematics Education' in The Handbook for Research in Mathematics Education Vol. 2 (Kluwer Academic Publishers, Dordrecht) (pp 827 878) WHEELER, D. (1974) R is/or Real (Open University Press, Milton Keynes)
Page 90
MIND THE 'GAPS': PRIMARY TEACHER TRAINEES' MATHEMATICS SUBJECT KNOWLEDGE

Tim Rowland, Caroline Heal, Patti Barber, Sarah Martyn Institute of Education, University of London Recent changes in the curriculum for Initial Teacher Training incorporate a stronger focus on trainees' subject knowledge. (DtEE, 1997) Some evidence would seem to support this shift of emphasis. In the US, Kennedy's research (1991) suggested that teachers' mathematical understanding is frequently limited, whilst in the UK Alexander et al (1992) called for improvement of the knowledge base of teachers in order to improve the teaching of mathematics. Inspection evidence identifies teachers' lack of subject knowledge and confidence in mathematics as being a contributory factor in low standards of mathematics attainment of pupils (Ofsted, 1994). Circular 10/97 (DtEE, 1997) sets out what is considered to be the "knowledge and understanding of mathematics that trainees need in order to underpin effective teaching of mathematics at primary level". From September 1998, audit and remediation of students' subject knowledge is statutory. All providers of ITT must audit trainees' knowledge and understanding of the mathematics contained in the National Curriculum programmes of study for mathematics at KS 1 and KS2, and that specified in paragraph 13 of this document. Where gaps in trainees' subject knowledge are identified, providers of ITT must make arrangements to ensure that trainees gain that knowledge during the course ... (DtEE, 1997, p. 27) In this paper, we describe our approach to the audit of the mathematics subject knowledge of a cohort of 154 trainees following a one-year primary PGCE course, and we offer some preliminary findings. Our research follows a number of different directions; here, we offer a (fore)taste of each of them.
APPROACH AND TIMING The structure of the primary PGCE at the Institute is perhaps unusual, in that the methods course for each core curriculum subject is taught in three intensive 5- or 6-day blocks, one in each term. The blocks for mathematics are timetabled first in each term, so that by the middle of January, with fully six months of the course remaining, the main content areas - number concepts and operations, data handling, mathematical processes, shape and space, measures, algebra, probability - have been 'covered' in lectures and workshops. This therefore seemed to us to be the optimum moment for an audit of subject knowledge, giving the trainees maximum opportunity and professional motivation to recall topics they had forgotten (for lack of use) since they did mathematics at school. A 1 Yz hour written assessment consisting of 16 test items in mathematics was therefore administered at this point of the course. Trainees had been given notice of the 'test' and a revision syllabus some six weeks earlier. Their response to each question included a self-assessment of their ability to complete it successfully. The scripts were marked and the response to each question coded either: secure, possibly secure, not secure. The middle category was created in recognition of the difficulty (resolved later) of making confident inferences from some of the written responses. Corresponding scores of 2, 1 and 0 respectively were recorded for each question and each student were entered on a spreadsheet. In midFebruary, an individual audit feedback sheet was returned to each student, with guidance (where appropriate) for further study.
Page 91
39 students (about 25% of the cohort) who had been found to be secure in 15 or more of the 16 topics audited were invited to become mathematics peer tutors. Following training for this task, they conducted one-to-one peer tutoring sessions with all other students (on average, three per peer tutor) in April, writing a feedback sheet on each of their tutees. TRAINEES' MATHEMATICAL THINKING One dimension of our research is to identify what mathematics (within the remit of Circular 10/97) primary trainees' find difficult, and the nature of their errors and misconceptions in theses areas. Facilities in the four 'easiest' and 'hardest' of the 16 items audited were as follows:
1 The
written response gives a high level of assurance of the knowledge being audited 2 A
secure response scores 2, which is therefore the maximum possible for the mean.
It is important to try to interpret some of these findings within the context of the course and the audit. For example, some standard errors and errors and misconceptions associated with ordering decimals are well-documented (Mason and Ruddock, 1986). For this very reason, these common difficulties are brought to the attention of trainees in the taught course, as an important detail of the professional knowledge they need to acquire. This may account for the subsequent low incidence of such mathematical errors in these trainees. For other topics, such as generalisation, the taught course appeared to have been less influential, or the topics themselves inherently more demanding. Others, such as Pythagoras' theorem, were not given detailed attention in the course. Ongoing analysis of responses to the more difficult items is uncovering errors which illustrate a continuum of 'gaps' in trainees' subject knowledge in particular areas. By way of illustration, the 26 insecure responses to the following question (on generalis at ion) were scrutinised.
Responses were assigned to categories named (for example): blank, irrelevant, incoherent, checks but does not generalise, makes false generalisation. The last of these was perhaps the most interesting. The response of one student most clearly represented the 7 trainees who made one particular error:
q2...
Page 92
These responses appear to focus on the first example (3+4+5=3x4) to the exclusion of the other two, or to indicate inability to see the second two examples as counterexamples to the proposed generalisation. The '3' after the '=' is not perceived as the one constant term common to the three equations. It is perhaps not difficult to share the concern of the TT A about prospective primary school teachers who, for example, find it so difficult to perceive and communicate a 'one-ness' of form (let alone of meaning) in the three equations. It is also questionable whether the IT A's fondness for self-study (DfEE, 1997, p.27) is misplaced in the face of such cognitive obstacles.
SUBJECT KNOWLEDGE AND CLASSROOM PERFORMANCE
Research at King's College into effective teachers of numeracy suggests that what matters is .. , not formal qualifications or the amount of formal subject knowledge, but the nature of the knowledge about the subject that teachers have. (Askew et aI, 1997, p. 93) With this in mind, another strand of our enquiry investigates whether a significant link between subject knowledge, as measured by the audit, and students' performances on teaching practice can be identified. Given the King's findings, it might be expected that trainees' teaching practice performance would be independent from their audit of 'formal subject knowledge.
Method
Students were assigned to a 2-way classification: I: Audit scores (maximum score being 32). Category A, B or C corresponding respectively to audit score above 30 (i.e. perfect or nearperfect), between 30 and 24, below 24 (of whom almost all were insecure on 3 or more items) II: Teaching practice performance Students were categorised as I (very strong/strong), 2 (capable) or 3 (weak) on the basis of a formative grade profIle given on their first (Spring) term teaching practice for planning, teaching and assessment. It should be noted that these grades - the best available to us at the time of writing do not relate solely to the teaching of mathematics. These data are shown in the contingency table below, together with expected frequencies (in brackets) based on the null hypothesis that audit performance and teaching performance are independent.
Page 93
A chi-square test can be applied to these grouped audit and teaching practice data. The contingency table has 4 degrees of freedom, and X2 = 5.59, with probability p = 0.23. Such a high probability supports the hypothesis that teaching practice performance and audit performance are independent. Some 'extreme' cells may seem to cast doubt on this conclusion. For example, the actual number of C3 students is 50% more than 'expected'. Is then a low audit score, in particular, a significant predictor of weak teaching performance? Again, the answer must be 'no'. On the hypothesis that the distribution of the 43 students with low audit scores across the three teaching grades is the same as that for the whole population, a binomial model B(43, 24/154) gives the probability of 10 or more category C3 trainees to be 0.12 i.e. not significant, even at the 10% level. We are currently revisiting these questions, updating our data with grades specifically related to the teaching of mathematics in the final (Summer) teaching practice. TRAINEES WITH SECURE MATHEMATICS SUBJECT KNOWLEDGE A third dimension of our research concerns the histories, attitudes and professional trajectories of trainees who score highly against our audit oftheir 'formal' mathematics subject knowledge. Sixteen students out of 154 were secure on all of the 16 audit items. 11 ofthese have A level mathematics (although another 16 with A level are represented throughout the 'top' two-thirds). Eight of the 16 have been selected for case study - four specialising in Early Years, four in Middle Years (KS2). Their degrees are in Psychology (4) English, Social Anthropology, Society and Technology, Linguistics and German. Four have A level mathematics. In the self-audit, all but one were confident about their own ability to teach the mathematics subject content of the audit to someone else. Here, we focus on the exception, an Early Years specialist whom we shall call Frances. A mature student, she has a 2.1 Cambridge degree in Social Anthropology, three A levels at A grade and a level grade 2 in mathematics. She is academically well qualified but not particularly so in mathematics. A tension between her pedagogical knowledge and beliefs, and the practices she encountered in school was apparent in her first teaching practice. The class teacher taught mainly by using worksheets and this was not what the student wanted for her own teaching. In a coursework essay, Frances reflected: 'In my own teaching practice there was a problem in laying the basis for progression. In the scheme of work agreed, I planned to teach the concept of difference, then show how it can be represented by the minus sign. However, when I team taught with the class teacher at the end of the practice, the teacher said she didn't like to teach subtraction as difference but only as 'take away' because it was confusing at this age.' Grossman et al (1989) believe the task of transforming disciplinary knowledge into content suitable for students is one of the central skills of teaching. Frances argues: 'However, if teachers' understanding is deeper and the learning objectives are related to how the subtraction sign fits into the whole "form of discourse" (Aubrey, 1994), it is relevant to introduce the idea that a minus sign can mean different things to quite small children, especially once they are familiar with more than one subtraction structure. I was consciously using my knowledge both about mathematics, its nature and its concepts and my knowledge and experience about how young children learn. Presenting young children with
Page 94
the concept of mathematical difference without linking it to their existing practical experience would have been akin to presenting them with a foreign language with no scope for translation.' This points to Frances' own understanding, and how a student's thinking may be more theoretically sophisticated than that of the class teacher with whom they are training. Frances was the only one of the group who was not confident about her ability to teach the subject matter of the audit, even though her 'test' result did indicate that her subject knowledge was highly secure. It is easier to make judgements about a lesson and how to extend children's understanding if the teacher is confident of their subject matter (Pollard et al, 1997). This factor underpinned a second tension for Frances, between what she believed to be best for pupils and her reluctance to implement it. Elsewhere in her writing, she recognises her lack of confidence 'To develop powers of reasoning, children need opportunities for mathematical investigation. In not capitalising on such opportunities, I failed to present mathematics as an investigative problem solving subject. This was an example of how use of subject knowledge can be affected by confidence and experience. At an intellectual level I appreciated the importance of mathematical investigation, but did not have the confidence to pursue it. Frances is eloquent and reflective about her practice. She is aware of why she had problems but needs the confidence of experience to pursue her own ideas and work upon these. She illustrates why a high level of subject knowledge and understanding of pedagogy are clearly not enough because her problem is a lack of confidence in herself. THE PEER TUTORING PROCESS The peer tutoring arrangements have been the focus of our fourth research initiative. 32 students who scored highly in the subject knowledge audit agreed to act as peer tutors to other students, following a briefing session. These met with their tutees one-to-one, following a prearranged schedule of appointments, and reviewed the troublesome items in their tutee's subject knowledge audit paper. Out of 32 peer tutors, 18 agreed to audio-tape record the sessions. Of these, 15 persuaded at least one of their tutees to be tape-recorded. As a result we have several hours of recorded material that we are in the process of transcribing for analysis. The peer-tutorials seemed highly-charged events for the participants. No doubt the negotiated audio taping contributed to this, but it seemed as if this extra dimension was insignificant compared with the self-consciousness engendered by the occasion itself. Most noticeable was the degree of controlled anxiety amongst the peer tutors. They felt their professional responsibilities keenly, notwithstanding their professed confidence; discussion with them on the day suggested that they had a strong sense that both their subject knowledge and their pedagogic skills were on the line. The material from the transcribed tape-recordings is very rich. The preliminary analysis is suggesting several themes that could be pursued. For example, the material could be explored in terms of the range of discussion taking place around particular questions on the audit. The hierarchy of difficulty of the questions is such that much of the taped material focuses on a fairly narrow range of the more difficult topics (see the earlier table). Clearly some work on the data related to each question will be rewarding. By contrast there is a much smaller amount of material relating to the easier questions in the audit, and these are interesting cases because they tend to involve, as peer tutees, those students whose subject knowledge is apparently most seriously deficient as judged by our audit.
Page 95
As teacher educators engaged on a day-to-day basis with making professional judgements about the effectiveness of the teaching of novice teachers, the most seductive aspects of a preliminary review of the material relates to the sense that some of the peer tutors are more effective teachers than others. This 'sense' requires a great deal of disciplined unpacking, but is informed, for example, by the contrast between the determination of some peer tutors to 'explain' and the willingness of others to listen. Follow-up work which invited peer tutors and/or tutees to evaluate the teaching and learning (with or without access to the taped record) would be an interesting project. In addition, issues of tutor/tutee gender relations, of patterns in turn-taking and the exploration of many language issues are all potentially available. One of the most interesting features of this initiative is its particular context as an example of peer tutoring in the context of learning to teach. There is a longstanding interest in peer tutoring (Topping, 1996) but little work has been done on its potential within teacher education. The relationship between subject knowledge and pedagogic knowledge is not a simple one (Shulman, 1987, McNamara, 1991), and through their participation in peer tutoring, students could be encouraged to confront and explore it. If this approach to addressing students' needs for enhanced subject knowledge proves fruitful, we could consider extending it through sensitive use of the taped material as part of the learning experience of the course. We need to form a view about whether these peer tutoring methods have the potential to support the remediation of deficiencies in subject knowledge and to develop teaching skills, and if they do, how to refine and develop them.
REFERENCES
Alexander, R; Rose, J & Woodhead, C (1992). Curriculum Organisation and Classroom Practice in Primary Schools. London: Her Majesty's Stationery Office. Askew, M; Brown, M; Rhodes, V; Johnson, D and Wiliam, D (1997). Effective Teachers of Numeracy. London: King's College. Aubrey, C. (ed.) (1994) The role of subject Knowledge in the Early Years of Schooling. London: Falmer. DillE (1997) Teaching: High Status, High Standards (Circular 10/97). London: Her Majesty's Stationery Office Grossman, P., Wilson. and Schulman, L. (1989) Teachers of Substance: Subject Matter Knowledge for Teaching. In Reynolds M. (ed.) Knowledge Basefor the Beginning Teacher. Oxford: Pergamon, for the American Association of Colleges for Teacher Education. Kennedy, M (1991). An agendafor research on teacher learning. NCTM Special Report. Michigan: Michigan State University. Mason, K. and Ruddock, G. (1986) Decimals. Windsor: APU/NFER-Nelson McNamara, D. (1991) 'Subject knowledge and its applications: problems and possibilities for teacher educators', Journal for Education for Teaching, 17 (2), pp. 113-28. Ofsted (1994). Science and mathematics in schools: a review. London: Her Majesty's Stationery Office. Pollard, A., Thiessen, D. and Filer, A..(ed.) (1997) Children and their Curriculum: the perspectives of primary and elementary school children. London: The Falmer Press Shulman, L. (1987) 'Knowledge and Teaching: Foundation of the New Reform, Harvard Educational Review, Vol 57, No 1. Topping, K. (1996) 'The effectiveness of peer tutoring in further and higher education: a typology and review of the literature', Higher Education, Vol 32, No 3, pp. 321-345.
Page 96
INFLUENCES ON STUDENT TEACHERS OF MA THEMA TICS D N (Jim) Smith Sheffield Hallam University This paper reviews work in progress on a study of the nature o.linfluences upon student teachers of secondary mathematics. A practitioner research model has been employed in three phases to date, involving three cohorts {istudent teachers on the one year post-graduate certificate in education course (PGCE) at Sheffield Hallam University. Thefindings suggest that mathematics teachers were motivated to advise student teachers most frequently about aspects of class management. Other aspects of guidance offered within this structure were mainly focused on explanation, examples and exercises. There was some attempt to exhort student teachers to use a range o.l activities, but little guidance was offered about specific activities.
Phase 1 1995/96 This phase is described in more detail by Smith (1996a). To summarise, many of Sheffield Hallam's student teachers had commented upon their perception of a strong difference between the views of school teaching propounded by the university based teacher trainers (e.g. Smith, I 996b) and those propounded by many teachers in their practice schools. This apparent dichotomy might be characterised as university tutors being seen to be advocating strong but idealised positions and approaches whilst the influential teachers in school are characterised as being more pragmatic and practical.
The aim of the first phase of the study was to explore and describe the gap between university ideals and school practice in terms of advocated mathematics teaching styles. To inquire into these questions a survey was undertaken of the views of all twenty-one students on Sheffield Hallam University 1995/96 cohort of mathematics PGCE students.
This survey confirmed that the students did perceive a gap between what was being advocated by university tutors and the practice observed in their placement schools;
These particular student teachers regarded school teachers of mathematics to be doing too much routine practice work and insufficient investigative work .
The student teachers appeared to be convinced of the desirability of engaging their pupils in active learning tasks (although one or two seem to be indicating that tutors were overstating the case in an effort to counterbalance school practices).
Many of the student teachers were aware of school teachers adopting different ways of working with different ability pupil groups.
These particular student teachers allied their own views closely to the perceived views of their university tutors.
It is reasonable to speculate that once these students become full time teachers, the initial influence of the university tutors may fade over the years. Phase I of the study had not been designed to identify any changes in views.
Phase 2 1996/97
This phase studied nine student teachers as they progressed through their PGCE year in 1996/97. The students completed surveys at three points in the year; on entry, half way through the year and during the final session at the end of the year. The surveys were used to identify which individual students were making changes in their views regarding teaching approaches. These individuals were subsequently interviewed to obtain more detail.
There were few generalities in the changes experienced by the students, the most significant being a convergence towards the view that lower attaining pupils would benefit from the use of practical equipment and practical tasks.
Most of the changes in views occurred at the individual level, where changes made by students appeared to follow a pragmatic line, responding to the classroom environment student teachers found in their practice. However, some students made changes in their teaching approaches happily and consciously, acknowledging their initial lack of background knowledge. Other students made such changes pragmatically, but sometimes reluctantly as the changes conflicted with strongly held views on the nature of the task. This bears comparison with other studies, e.g. a study of student teachers' implementation of investigative approaches (Cooper, B. 1990) where strongly held views about the correct nature of mathematical education persisted through the PGCE year.
A significant issue that emerged during phase 2 was that of the relationship between the student teacher and the various classroom teachers from whom they were temporarily taking over. The
Page 98
particular aspect that caused difficulty was the student teachers frequent attempts to emulate the class teachers' style and approach. At best this can be an emulation of a range of effective styles, at worst the student is trying to emulate a range of conflicting styles and sometimes attempting to do so with a single class.
Phase 3 1997/98
This consisted of an investigation into the influence of mathematics teachers on student teachers of secondary mathematics, with a particular focus on the selection of pupil activities in the construction of lesson plans. Evidence was collected in three ways;
I. a documentary analysis of lesson observation feedback forms (10 per student) which are used to provide the 14 student teachers with written guidance on their teaching 2. a questionnaire survey completed by 26 out of 61 mathematics teachers and school-based mentors who were offering the student teachers advice 3. interviews in which the 14 student teachers lesson plans were used to stimulate recall of planning decisions regarding the selection of pupil activities
The documentary analysis of lesson observation forms was focused specifically on advice to the student teacher about the choice of tasks for pupils. In summary;
The advice relating to teaching approaches was predominantly concerned with exposition, examples and exercises. Students were advised about the clarity of their exposition, both orally and in relation to board work. The quantity, quality and graduation of examples and exercises were commented upon. One third of the teaching practices passed with no other aspects of pupil tasks being commented upon, that is no comment on alternate approaches involving pupil discussion of mathematics, investigative work, practical work, or non- routine problem solving. (Cockcroft, 1982, para. 243) . In general there was little comment on pupil learning activities except some requests for 'more variety', but no alternate tasks were offered. It is possible that summative lesson observation forms are not an appropriate place for teachers to advise student teachers on alternative approaches, but that alternatives might well be considered before rather than after teaching.
Page 99
Each of the student teachers in the cohort worked with a number of classes in their placement schools during the first half ()f the year. The mathematics teachers who normally took these classes were asked in a questionnaire to idcntify the range of advice that they had provided to the student teachers. The total number of mathematics teachers invited to participate was 61. The total number of respondents was 26. Some 13 of these respondents had received training as school-based mentors. The questionnaires for teachers raised a widcr range of issues than the documentary analysis, including advice in relation to pupil tasks but also questions relating to the relationship between student teacher and class teacher and questions relating to advice on class management.
I
'!
Some differences were identified between school-based mentors and class teachers. In comparison with class teachers, mentors were more strongly incl ined to identify student teachers strengths more likely to offer student teachers challenges always involved in the assessment of student teachers more likely to see themselves as guides to the school more likely to see themselves in a coaching role with the student teacher more strongly identified with a reflective practitioner model of development more likely to attempt to attempt to exemplify good practice more involved in the planning of the student teachers learning experiences
These differences could be thought surprising. The very limited time involved in school-based mentor training cannot be expected to make a great impact in comparison to the long term experiences of a professional career. However, school-based mentors are different from other mathematics teachers in the sense of generally having opted into the teacher training role and in tending to be more senior members of staff. The questionnaire responses regarding advice offered to student teachers can be summarised as follows;
Class management was the main focus, with frequent advice about keeping pupils busy.
100
Page 100
The second most frequent advice related to pitch and differentiation; again this was often phrased in terms of keeping the pupils busy. In other words, the most able pupils should not be allowed to run out of work, and the least able pupils should have work which is accessible and keeps them occupied. Advice about choice of activities was not a particularly strong feature but where advice was to be offered, it would mainly involve aspects of explanation, examples and exercises. There were some appeals for variety without specific alternatives being suggested.
The student teachers were interviewed about their lesson planning decisions, with the intention of identifying the sources of ideas for pupil activities. On a straightforward count of identified influences, the major sources of pupil activities in order of frequency were:
I. the student teachers "own ideas" (57%)
2. ideas developed from the pupil textbook (20%) 3. mathematics department teacher / school-based mentor (9%) 4. university sessions (8%) 5. other i.e. peers, in - service course, reference texts (6%)
Some conclusions
The evidence accumulated from three different approaches used in phase 3, together with evidence from phases I and 2 suggested that mathematics teachers are motivated to advise student teachers most frequently about aspects of class management. This may encourage student teachers to view class management and the learning of mathematics as though they are entirely separate issues. Other aspects of craft knowledge were also shared, but were subordinate to a concern with class management. The 'other aspects' of guidance offered in this context were mainly focused on explanation, examples and exercises. There was some attempt to exhort student teachers to use a range of activities, but little guidance was offered about specific activities.
Student teachers in this study have relied predominantly on their own devices in order to produce pupil activities, suggesting that direct influences on their choice of activities are weak. It may be that external influences remain, but are influences on the type of activity undertaken rather than the particular activity. This might provide an opportunity for further research.
10\
Page 101
Suggestions for" Further Resear"ch

With a lack of direction from school on the issue and a relatively limited quantity of suggestions from HE, student teachers generally have chosen to devise their own pupil activities. This raises for me the related issues of what the student teachers are trying to achieve with their activities and how they design the activities. Clearly there are obvious targets, such as the teaching of lesson objectives. What kinds of activities do they design to meet these targets? What principles, in addition to the maintenance of order, are being used to choose between competing alternatives? How do student teachers articulate these principles? Where do the principles originate? Given an ambitious suggested pupil activity, how do student teachers modify it to make it more "workable" for them in the classroom. What is the thinking behind such modifications?
Acknowledgements
I acknowledge with thanks the support and guidance of Peter Gates (Nottingham University) and the critically constructive comments of Brian Hudson (Sheffield Hallam University).
References
Cockcroft, W. H., 1982, "Mathematics Counts", HMSO Cooper, B., 1990, "PGCE Students and Investigational Approaches in Secondary Maths", Research Papers in Education, Vol. 5, NO.2. Smith, D. N., 1996a, "College Ideals and School Practice ", Mathematics Education Review no.7, January 1996, pp 25-30. Smith, Jim., 1996b, "Getting Started", Mathematical Association, April 1996, 92pp, ISBN 0 906 588 340
10'1-
Page 102
Possibilities in Peirce's Existential Graphs for Logic Education

Adam Vile and Simon Polovina GR&ND South Bank University
In our experience students find learning logic difficult, this view is supported in the literature. Charles Sanders Peirce, logician, semiotic ian, teacher, envisaged a representation that would provide tools to enable everyone to reason with formal logic. To this end, and basing his work on his own semiotic principles, he developed a system of graphical reasoning. In this session we will present Peirce existential graphs and consider this form of reasoning as providing possibilities for improving logic teaching and learning. We will also outline our plans for a future study to evaluate the Percian approach.
Logic, Mathematics, Education

"Now although a man needs not the theory of a method in order to apply it as it has been applied already, yet in order to adapt to his own science the method of another with which he is less familiar, and to properly modify it so as to suit it to its new use, an acquaintance with the principles upon which it depends will be of the greatest benefit. For that sort of work a man needs to be more than a mere specialist; he needs such a general training of his mind, and such knowledge as shall show him how to make his powers most effective in a new direction. That knowledge is Logic. "(Peirce 1931-1958, 2.67 ) "much of the area of logic is not easy to comprehend ,", with practice and concentrated effort one can achieve a useful understanding" (Kelly 1997) Logic is the backbone of mathematics and computer science, yet it is clear from the limited research in this area that many intending mathematicians and computer scientists have very limited logical facility. For example the recent London Mathematical Society report (Howson 1995) suggests that students ability to prove is very poor and Barnard (1995) has shown convincingly that simply negating a statement can be problematic. Of course logic is not mathematics, and logic education is not mathematics education yet there are such strong connections between logic and mathematics that we feel it appropriate to discuss logic education at a mathematics education conference. Logic has a place in the history of mathematical foundations and in mathematical notation, and today we find that language of mathematical logic from the earliest mathematics classrooms. Gibson (1986) advocates teaching logic at school, and even wants to replace mathematics with logic at this level. His reasons are twofold, the first that logic is more useful than mathematics in everyday terms and the second that anything that is of value in mathematics terms of education is also in logic. We cannot deny the importance of logic in our technologically driven daily lives.Ifwe accept Gibsons' idea however, teaching logic remains beset with problems. Barnard (1995) found, after testing 684 students across all age ranges for their ability to negate statements, that in general whatever the level of formal training students are not very good at such negation. Common errors were the negation of the main verb only (stay awake became fall asleep) and incorrect negation of the quantifier
(there exists became there does not

10 3
exist). Interestingly he also found that students were influenced by context, those less contextualised questions were answered much better that those that actually made some sense. We can add, along with other teachers of elementary logic, a wealth of anecdotal evidence to support the claim that there are problems in learning logic. As one example consider the problem of material implication through the "raining" context (box 1) which always causes our students intense problems in terms ofthe truth table representation and the case in which it doesn't and I don't bring my umbrella.No textbook that we have seen explains this adequately. Confusion is compounded by statements such as:
"One could disagree with the reasoning behind this interpretation but we will avoid philosophical tangles and simply remark that this definition provides a coherent and fruitful element in the formalisation of classical logic" (Kelly, 1997 p.7).
Representations
The reasons for these problems are as yet unknown although there is evidence to suggest that the degree of symbolism and abstraction in logic (Dubinsky, Elterman et al. 1996) and in mathematics in general (Vile and Lerman 1996) cause conceptual difficulties. The problems that Barnard observed with context point towards language, and other choices of representation as problematic and ambiguous. This notion is not new and the philosopher, semiotician and logician C. S. Peirce, as early as 1906, suggested that a more graphical approach to logical reasoning would help to clarify the process. To this end he developed existential graphs which are a graphical representation of predicate logic with the express purpose of logic teaching and clarity in mind (Houser, Roberts et al. 1997). Peirce expressed is view clearly during one of his lectures on logic:
Our purpose, then, is to study the workings of necessary inference. What we want, in order to do this, is a method of representing diagrammatically any possible set of premisses, this diagram to be such that we can observe the transformation of these premisses into the conclusion by a series of steps each of the utmost possible simplicity. What we have to do, therefore, is to form a perfectly consistent method of expressing any assertion diagrammatically. The diagram must then evidently be something that we can see and contemplate. (Peirce 1931-1958, 4.429-430)
Peirce apparently always thought in pictures, he attributed this to his left-handedness. He combined his Pragmatism, semiotics and logic into a clear system of diagrammatic thinking that would provide diagrams upon which one could experiment (4.530), much like the way scientists use diagrams in their own work. Traditionally logicians have been critical of the use of diagrams as an aid to logical deduction but Barwise and Etchemendy (1996) support Peirce's view and challenge
the "logocentralism" of logic, arguing for the use of diagrams to support reasoning on the grounds that they contain much more knowledge that equivalent forms. There has been much recent interest in applications of Peirce logic (Lukose, Delugach et al. 1997). Peirce suggested that" [ through a] System of diagrammatization .,. any course of thought can be represented with exactitude" (1906, p492) and it is well known, in mathematics at least, that diagrammatic representations aid comprehension (Pimm 1996). Allwein and Barwise (1996) demonstrate the power of a graphical environment in logical reasoning and results in teaching seem positive, although (to our knowledge) this has not been fully investigated through empirical research. It would seem that there is general support for the view that a less abstract approach to logical reasoning would provide possible route to overcoming the identified problems in learning logic. In this short paper we will take a look at the expressive power of Peirce graphs with a view to demonstrating their possibilities for logic teaching
Peirce graphs, a primer

We want to suggest that existential graphs are more visual, more holistic, and more transparent to the underlying logic that traditional symbolic representations. Here we will give a simple introduction to Peirce existential graphs, for more details see, (Hammer 1996) or (Polovina and Heaton 1992). Existential graphs consist of two elements a sheet of assertions, which is essentially a blank sheet on to which assertions may be placed, and a cut which is a ring around an assertion equivalent to ~( ... ), thereby describing the 'not' relationship, also known as negative contexts. Combining assertions and two negative contexts we can form a scroll, a basic if then statement. Figure 1 shows a scroll representing p => q .
The mere co-existence of each propositional element in existential graphs automatically describes the 'AND' relationship Take for example Modus Ponens, represented in traditional propositional logic as
(p
(p
=> q)) => q. Existential graphs always represent logic denoted in 'and' and 'not' form, thus
the antecedents of Modus Ponens can be represented by the first of the graphs in figure 2. According to Peirce's rules of existential graphs' inference, any dominated element that matches a dominating element may be 'rubbed out' or deiterated. As p is dominated by its matching p in the outermost context, it can be deiterated leaving q surrounded by two rings alone. After the removal of the constraining p within the single negative context level, the resulting graph simply states
ID
Page S 105
- (-q) which, by double negation, results in q being asserted. We can therefore describe Modus Ponens by the following existential graphs:
Fig 2 From this example we can begin to see that existential graphs vividly illustrate the contextual interrelationships between the propositional. The above existential graphs reveal that q is asserted once p exists, since q exists in the context of p. Apart from avoiding the need to learn rote 'rules of inference' from the outset, the visuality that existential graphs offer obviate the requirement to consider truth tables. We can see from the examples thus far that the conjunction of all the graphs, which is automatic as explained earlier, results in true. This consistency is evidently not the case in the following, as deiteration and double negation would indeed confirm:
Clearly p and ~q, as well as ~p and q, are inconsistent. Thereby (-q n (p ~ q)) ~ p is false. Similarly (p
n (p ~ q)) ~ -q is also false. All this can be seen to be so without the learner needing to memorise or
draw up truth tables, as we have observed, this is a difficult task for students of logic. Of course; we could devise more complex examples that would highlight these advantages. For instance could the learner better tackle problems based around the following interrelationships if presented in traditional propositional logic form?
10(0
Page 106
As we would expect - x would result in the deiteration of x and its negative context, thus enabling y to be double negated to release y into the outermost context and thereby be asserted. This also shows us that -x ~ y and, for that matter, -y ~ x. However what would x give us? After deiteration the result would be (fig 5):
Here y cannot be determined from x as such, though we have shown that no inconsistency arises. As it happens there is a further existential graphs rule that can brought into play. This additional rule states that anything that shares the same context with an empty negative context, known as the 'empty clause', as in the above, can be removed. This is because logically the truth of any context which contains the empty clause is always false. The result is (fig 6):
The double negations cancel out simply to true. This becomes useful when faced with a problem such as given by the following example (fig 7):
Value of Peirce Graphs

The ability to make sense of existential graphs depends on connections with signs and not objects, there are certain spatial effects such as adjacency and inclusion that make the relationships and the
structure more transparent. Peirce Graphs offer a clear, graphical, unambiguous way to approach logic teaching. Logic students need only to understand a small number of simple rules and with experience may be able to use the diagrammatic representation to their advantage. No doubt this simplicity is what Peirce would have intended. However there are a number of problems with this form of representation, none least the degree of complexity that can arise with problems of high order (Polovina and Vile 1997). This form of logic does have its limits, nevertheless it was not intended as a replacement for symbolic logic, merely as a teaching and reasoning tool. Existential graphs seem so good at explicating logical processes that they also lay open these processes to critique in themselves, showing the limits of first and second order 'crisp' logic in support of reasoning and complex, real world decision support (Polovina and Vile 1997) Some may be critical of existential graphs on the grounds that it is not a usual notation and is therefore not at all useful for intending computer scientists and mathematicians; additionally it becomes very complex in its beta form for first order predicate logic. As such it could be argued that it is only pedagogically valuable in its alpha representation, as a middle step towards more formal (yet constraining?) representations.
Conclusions and the future
We hope that in this short exploratory paper we have been able to illustrate the possible value for logic education of conceptual graphs. Within our experience of teaching logic to first year undergraduate computer scientists we have come upon a number of recurring problems. The majority of these we believe are caused by inability of students to make appropriate sense of the symbolism, and its abstract nature, which in keynote cases (in terms of development in the learning of logic concepts) refuses to correspond with intuition. In other areas of mathematics education this phenomenon has been, and continues to be, examined in detail (for example see Pimm (1996)). Conclusions indicate the possible presence in individuals of cognitive gaps (Herscovics and Linchevski 1994) which divide senseful symbol use from incomprehensibility, and notions of reification of signs so that they become treated and manipulated as if they were objects. We want to suggest that problems with symbols and symbolic representations arise from their inherent opacity, and we would argue that initial introductions to mathematical subjects be as concrete as possible. Existential graphs are a good example of a less semiotically demanding representation for complex linguistic and logical structures. We hypothesise that Peirce logic is a better first approach for logic teaching that any other, and should replace the truth table representation which is the common first port of call. We intend to test this hypothesis with a comparative study during the next academic year and would be happy to
Page 108
report our results at this conference. If our hypothesis proves true then we will develop further dynamic teaching resources and use the evidence to institute a change in our own institution, which may possibly be followed by others. Peirce "would be the very first to insist that logic cannot be learnt from logic books or logic lectures" (Houser and Klosell992) and he remarks that, although it is useful, a 'mathematical head' is not necessary for the study of logic. To students of our University this would be a great relief. We suggest and hope to establish through research that this is only so if we remove the strain of making sense of formal logic symbolism and replace, or at least precede, it with the more transparent Peircian representation - existential graphs. Peirce had only a short time as a lecturer in logic, not enough to put his ideas to good use in education. We hope to re-open his logic casebook and begin again the clear explication of logic that he began as far back as 1882.
References Allwein, G. and J. Barwise, Eds. (1996). Logical Reasoning With Diagrams. Studies in Logic and Computation. Oxford, Oxford University Press. Barnard, T. (1995). The impact of meaning on students ability to negate statements. Proceedings of the ninteenth International Conference for the Psychology of Mathematics Education. Recife, Brazil. 2: 30-10. Barwise, J. and J. Etchemendy (1996). Visual Information and Valid Reasoning. Logical Reasoning With Diagrams. G. Allwein and J. Barwise. Oxford, Oxford University Press. Dubinsky, E., Elterman, et al. (1996). "The Students Construction of Quantification." For the learning of Mathematics 8(2): 44-51. Gibson, R. (1986). "Logic as a Core Curriculum Subject: its case as an alternative to mathematics." Journal of Philosophy of Edcation 20(1): 21-37. Hammer, E. (1996). Peircian Graphs for Propositional Logic. Logical Reasoning With Diagrams. G. Allwein and J. Barwise. Oxford, Oxford University Press: 129-143. Heaton, J. E. (1994). Goal Driven Theorem Proving Using Conceptual Graphs and Peirce Logic (PhD Thesis). Department of Computer Studies, Loughborough University of Technology. Herscovics, N. and L. Linchevski (1994). "A Cognitive Gap Between Arithmetic and Algebra." Educational Studies in Mathematics. 27(57-64). Houser, N. and C. Klosel, Eds. (1992). The Essential Peirce: Selected Philosophical Writings, (1867-1893). Bloomington and Indianapolis, Indiana University Press. Houser, N. a., D. Roberts, et aI., Eds. (1997). Studies in the logic ofC.S.Peirce. Bloomington, Indiana University Press. Howson, G. (1995). Tackling the Mathematics Problem. London, London Mathematical Society Kelly, J. (1997). The Essence of Logic. London, Prentice Hall. Lukose, D., H. Delugach, et aI., Eds. (1997). Conceptual Structures: Fulfilling Peirce's Dream. Lecture Notes in Artificial Intelligence. Berlin, Springer-Verlag. Peirce, C. S. (1931-1958). Collected Papers of Charles Sanders Peirce. Cambridge MA, Harvard University Press. Pimm, D. (1996). Symbols and Meanings in School Mathematics. London and New York, Routledge. Vile, A. and S. Lerman (1996). Semiotics as a Description in Mathematical Domains. Proceedings of the 20th conference of the International group for the Psychology of Mathematics Education. Valencia. 4: 395402.
\0'1
Page 109
11 0
Page 110
What can be learnt by selecting anecdotes from a range of data? Exemplifying "noteworthy" mathematics with a small number of examples.
Anne Watson, University of Oxford Department of Educational Studies Abstract Analysis of data collected for qualitative research purposescan often generate further research questions, particularly if an open-minded approach is taken. In this paper I examine the use of anecdotes for a purpose other than that for which they were collected. I suggest that this is a justifiable research procedure so long as the questions asked of the anecdotes are appropriate, and the use of the results is sensitive to the mode of collection. Introduction This paper is about the validity of selecting and analysing anecdotes from a larger body of data as
an extension of the analysis of a larger study. I discuss how and why the anecdotes were selected, and what could be deduced as a result.
The research context The purpose of the main research project was to develop a grounded theory [Glaser & Strauss,
1967] of teachers' informal assessment practices in mathematics (Watson, 1995]. Data from the whole project included pre-interview observation notes, audiotapes of unstructured interviews with thirty teachers about informal assessment practices, longer term observation of ten pupils in two teachers' classrooms and observations of seven in-school moderation meetings.
Teachers were in their normal settings, their own classrooms or staffrooms, aware that they were the focus for research on assessment but were not aware of how the detail was to be used. They will have been more aware of assessment issues than usual because they were being observed or questioned about them. They had not deviated far from their normal practices of informal assessment which depended on the accumulation and development of impressions, complemented in general by accumulation of written artefacts or test scores. These practices are to be seen in the light of recent national training in assessment techniques within a national framework which includes mathematical thinking and processes among its descriptors.
Page 111
"\
The main body of data Interview data often included illustrative anecdotes about individual pupils, or pupils in general. Observation notes often included records of incidents which might then discussed with teachers. The long-term observations included stories about individual pupils over a term. Teachers were from upper primary and lower secondary classes, but the illustrative stories they used sometimes came from a wider range of their experience.
There are therefore several different sources of incident in the whole body of data: researcher's observations, teachers' recollections, discussion about observed incidents, written or taped records, and combinations of all these. Incidents which come solely from interview data have already been selected by the teacher to illustrate a point the teacher wished to make; incidents which come from researcher-initiated dialogue, or researcher's observations, are also necessarily selected at the datagenerating stage, but the researcher has a responsibility to examine carefully the purposes and criteria for selection. The method of data-collection recommended by Glaser and Strauss fop cit.] is to recognise that theory is being developed alongside data collection and deliberately to collect the kind of data that will illuminate, expand or refute the emerging ideas. Data collected by a researcher working alone may be particularly susceptible to subconscious editing, so the grounds for all editing need to be made explicit. This research aimed to produce as full a picture as possible of informal practices, and then to expose differences within such practices. Later stages were designed to focus more closely on particular aspects of practice which had emerged as problematic, and to raise questions about their operation. The researcher, therefore, selected incidents to observe, record and discuss which either confirmed practice represented elsewhere in the data, or offered other aspects of practice not yet represented, or which seemed to contradict what had already been learnt. There was a particular interest in data which might relate to justice and fairness in decision making about pupils. The incidents During the subsequent analysis and synthesis of a theory of the social justice of teachers' informal assessment practices it became apparent that teachers' decision-making about some isolated incidents did not always follow the patterns predicted in literature about interpersonal judgement. To explore this further within the existing data it was decided to extract incidents of teacherdecision-making from the data in order to begin identification of contributory factors. 'Sameness' and 'difference' between incidents would be explored, and 'sameness' and 'difference' relating to the literature would also be explored. It was hoped that what
112..
would emerge from these two comparisons would be a picture of what the literature did, and did not, adequately explain. The purpose of the selection and discussion of the incidents is therefore to identify, describe and contrast some classroom incidents during which, in some way, exceptional mathematical performance is displayed which the teacher notices and takes into account in her judgements. Incidents are selected which display features which exemplify or counter-exemplify the influences suggested in the literature and in earlier stages of the research. Features of these incidents can be compared to each other and related to outcomes in terms of teachers' judgements, in order to indicate answers to the questions: What factors influence the teacher to accept a single exceptional mathematical performance as noteworthy evidence to add to her picture of the student's general achievement, and what lead her to dismiss the performance as adding nothing of worth to the picture? Performance is taken to mean oral, written or physical communication which indicates the results of some mathematical thinking, such as giving answers, making suggestions or indicating something with a diagram. Exceptional means different to some norm, either for the class, or for one pupil, or relative to the teacher's expectations. Noteworthy means that it was recorded in some way, either mentally or in a formal way, and hence might contribute to the teacher's view of the pupil. Sample size The selection of anecdotes was limited in various ways: to what was already in the data; by the availability of stories which illustrated adequately the kind of incident required for the research; by the need to meaningfully compare incidents, i.e. they would need to have some common aspects in order to generate questions about their uncommon aspects; to those which related to the literature and previous research; to incidents the researcher knew enough about to describe meaningfully.
It must be remembered that the focus for this small study grew at the analysis stage, and hence the data had not been collected with this issue in mind. Most of the research had focused on the accumulation of judgements about pupils, not on judgements made on the basis of one incident. Hence it was not a surprise that very few suitable incidents were
Page 113
found in the data, and most of those were of students who were initially thought to be weak. Nevertheless it was felt sensible to proceed for two reasons: first, because no other work research has been found on this issue; second, because the aim is to indicate possible answers to questions, not to make generalisations. Given that even a single incident can generate doubts and research questions, it is therefore valid to have a small sample with which to work for the same purpose. The frame of reference for analysis Since the incidents have been selected for a particular purpose their use must be restricted to that purpose, otherwise different selection criteria might have been applied. They cannot, for instance, be used to exemplify general behaviour, for they have been selected to display contrasts, not similarities. They certainly cannot be said to say anything definite about the whole field, for the original data from which they were selected has also been subject to nonrandom selection. Hence one has to be very clear about what can, and what cannot, be usefully found in the incidents.
Another purpose of selection was to relate real incidents to the literature and research, so the incidents were searched for features which related to theory and also for common features which were not related to theory. Briefly, it had been suggested that in order to be thought of as strong (or weak) mathematicians students need to produce certain kinds of positive (or negative) learning behaviour [Eisenberg, 1991] to produce text which the teacher finds acceptable (or which differs from the teacher's desired type) [Morgan, 1996] to respond appropriately (or not) to the teacher's view of mathematics [Brown, 1995] to make a good (or poor) first impression [Nisbett and Ross, 1980].
The teacher then selects incidents to remember which support her view exaggerates them in her perceptions of the student ignores contradictory evidence sees the student's work in a positive (or negative) light [Nash, 1976] is reluctant to change her mind. Much of the literature on interpersonal judgement derives from laboratory psychology tests and may well omit certain situationally-specific features. Some studies suggest that time pressure and the importance of the decision also have some bearing on judgement. Use of
l\l t
anecdotes could, therefore, give clues about further factors which occur in non-experimental situations. Information about whether and how the incidents affected the teacher's decision and how the pupil's mathematical behaviour related to the rest of the group, the teacher's expectations in general and the pupil's past mathematical behaviour was extracted. The features so found are summarised in Table 1 which, while heavily reductionist in nature, nevertheless shows that application of theories of judgement to individual incidents fails to fully explain teachers' use of incidents in decision-making. Search for other common features, not expected from the literature, in incidents which produced similar outcomes can be commenced, and this would only result in suggestions for further research.
The incidents themselves and a much fuller analysis can be found in Watson [1998]. This is not the place to go into the content of the research further. The purpose is to show how a relatively small number of carefully selected anecdotes, while showing some support for theory, can reveal omissions in existing literature and hence raise research questions, or at least raise doubts about applicability of general literature to all settings. Anecdotes collected for other purposes can, with care and appropriate analytical methods, be used for other purposes.
Page 115
References Brown, L. (1995). The Influence of Teachers on Children's Images of Mathematics. In D. Carraher & L. Meira (Eds.), Proceedings of the 19th Conference of the International Group for the Psychology of Mathematics Education (Vol. 2, pp. 146-153). Recife, Brazil. Eisenberg,T. (1991). On Building Self-confidence in Mathematics. In F. Furinghetti (Ed.) Proceedings of the 15th Conference of the International Group for the Psychology of Mathematics Education (V 01. 2, pp. 6-16). Assisi, I tal y. Glaser, B.G. and Strauss, A.L. (1967): The Discovery of Grounded Theory: strategies for qualitative research. Chicago, Aldine. Morgan,C. (1996). Language and Assessment Issues in Mathematics Education. In L. Puig & A. Guttierez (Eds.), Proceedings of the 20th Conference of the International Group for the Psychology of Mathematics Education (Vol. 4, pp. 19-26), Valencia, Spain. Nash, R. (1976) TeacherExpectationsandPupilLearning . London, RKP. Nisbett,R. and Ross,L. (1980) Human Inference: strategies and shortcomings of social judgement. Englewood Cliffs, Prentice-Hall. Watson, A. (1995): 'Evidence for Pupils' Mathematical Achievements' For the Learning of Mathematics, Vol. 15,pp. 16-21. Watson, A. (1998, forthcoming):'What makes mathematical performance noteworthy in informal teacher assessment' Proceedings of the 2200 Conference of the International Group forPsychology of Mathematics Education. University of Stellenbosch, Cape Town
lI b From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 116
TEACHER TRAINEE STUDENTS' UNDERSTANDING OF OPERATION SIGNS David Womack University of Manchester Abstract Following the invention of operation signs by young children working m a tramj'ormationallyfocussed problem scenario, a similar transfonnational model was presented to 4th year teacher trainee students in which they also were asked to invent signs. The signs referred to the strategies of counting-on, counting-back and counting-up. The students' responses and their understanding of the difference between counting-back and counting-up will be discussed, and participants invited to give their own opinions on the mathematical status of such 'counting signs '.
Introduction:
The essential role of symbols in the development of mathematics in general and arithmetic in particular is unquestionable (Kaput, 1987, Sfard, 1994, Pratt & Garton, 1993). In the classroom, collections of objects are symbolised by number symbols which are manipulated according to rules. Because these rules correspond to how collections of objects behave in the reality, predictions can be made about the size of collections when these collections are added-to or taken-from etc. For example, 17 - 3 can describe the action of taken 3 objects from 17 objects. Augmenting this expression with the expression
= 14 displays the result of that action. When this result is omitted, as in '17 - 3 =' , is conventionally taken as a question (17 - 3 = ?). In the context of school
computations, this representedsituation can be realised or re-invoked by the child using concrete objects to model the expression and in this way, the result can be found experimentally. In this case the expression 17 - 3 =? takes the form of an instruction to perform the action necessary to find the answer. This dual function of description and instruction causes problems with the interpretaion of signs as the following investigation shows.
Background to the investigation In an earlier experimental teaching situation dealing with stepping stones (Womack, 1998),5 year-old
children carried out instructions and answered questions within a transformational framework. Results showed that they could confidently use signs which indicated how far they were to count-on or count-back along sequences of numbers. In a slightly different context they could also count-up the number of steps between two given letters or numbers (See Figure 1).
Counting procedures are not normally taught explicitly but are nevertheless necessary In order to compute mentally. It was of interest to know how far the intuitive skills of young children could be made explicit to teacher-trainee students when both types of sign were used in the same problemcontext. The signs invented by the children (Note 1) in the earlier
\\7 Page 117
experiment (see above) were sloping arrows which instructed the sign-user to step-on (or count-on) and step-back (or count-back). A different arrow sign was used to indicate that the number of steps between two stones was to be counted.
With the 5 year old children, the use of this latter counting-up sign was NOT used at the same time as the counting-back sign (to indicate counting-back was to be carried out), since it was felt there was a possibility for confusion. The stepping stone model used with children was initially based on the standard Oksapmin numeration system (Note 2) outlined by Saxe (1982), in which to count, one begins with the thumb on one hand and enumerates 27 places around the upper periphery of the body, ending on the little finger of the opposite hand. In the investigation described here, a letter sequence was used, to avoid possible confusion of numbers representing both positions and movements. The isomorphism between movement from body position to body position and movement from one letter to another is shown in Figure 2.
\\~
Page 118
The Investigation The following question sheet (Figure 3) using an alphabetic sequence model was given to nine 4th year teacher trainee students for completion in about ten minutes time in an 'open class' situation .
The results from all the students were then discussed in an open forum. A further question (Question 5) asked for an explicit symbolic formulation for these four broad questions. This required an expression in which the unknown (asked for) term was expressed explicitly by means of a relation sign linking the two given terms (letters in this case). Summary of Results These are shown in Figure 4. The most common response of the students was to use an arrow. However, in some cases the arrow pointed from one letter to another - indicating that the number of jumps between the letters had to be found; in other cases it pointed from a letter to a number, instructing the reader to count on that many jumps from the given letter. (In these cases, the question asked was assumed to be, 'Where does such an action take you?') Note also that in Question 4, the order of letters given in the written question is the reverse of their order in the alphabet. Questions 1 and 2: In question 1, all students drew an arrow from a letter, indicating perhaps that their thinking began with a letter. Student 5 began with a letter, but began the arrow from the number (i.e. from D, 7 forward). All students used arrows as signs. 4 students (nos. 1,2,3 & 4) used signs and number in parallel. Some used a box instead of the equals sign. 1 student used the sign after the number. 3 students used the sign before the number. 1 student used conventional signs before the number. Question 2: The arrow in each case indicated the direction of the movement along the letter sequence and students 1 to 4 indicated the 'missing' letters by a square placed in its correct
..
Page 119
sequential position (to the left of D).Their representation therefore referred to the alphabet model rather than the written words. Students 6, 7 and 8 seemed to be trying to say 'From D jump-back 3' using symbols in the same order as the written words.
Questions 3 & 4: 2 students (nos. 1 & 2) recognised that this was the converse question to that in
Questions 1 and 2. Therefore they put arrow and missing 'jump-on/back' box again in parallel. 3 students (5, 6 & 8) used the same form of arrow sign to mean both 'jumpon/back' (in Questions 1 and 2) and 'jump-on/back to' (in Questions 3 and 4). 1 student didn't make clear what the question was. 1 student used conventional signs - but not for conventional aggregation but to indicate the number of jumps to be made.
Question 4: 2 students (3 & 4) used the same logical system as students 1 & 2, but the expression
required reading in the reverse order (B/G). Their arrows indicated the relative position of the given questions.
Question 5: Attempts to pose Qs 3,4 'explicitly' was not achieved successfully except by giving
signs a dual interpretation. That is, using the arrow to mean 'jump-on 7' as in 'D ---> 7, but also to mean 'jump-on TO 7' (E ---> K.) Only students 7, 8 & 9 attempted to make the unknown the subject of the expression. Student 7 answered questions 1 & 2 explicitly, but not Questions 3 & 4. Student 8 answered Questions 1, 2, 3 & 4 explicitly, but used the same sign in all questions.
Page 120
Summary of student views

In discussion, students were uncertain whether they were using the same mental process to find a letter and to find a number. Therefore, if letters were to be replaced by numbers - as in the conventional counting procedure, clearly there would be a great potential for confusion. Some students argued that similar signs could be used to refer to similar processes (such as jump-on and jump-back) but should not be used for different processes. Others argued that two different mental processes might be represented by the same sign, provided the difference was understood from the context.
Discussion of explicit and implicit signs

The sign required for questions 1 and 2 represents an action necessary to perform in order to answer the question 'Which letter do you reach?' In conventional arithmetic this would correspond to the 'take-away' sign in the conventional expression 17 - 3. However, the sign required for questions 3 and 4 is a meta-sign, in that its function is to ask the reader to find another sign. This meta-sign represents the action required to find another action - that needed to move from one given letter to another letter - a number of jumps. For example, in conventional arithmetic, given an initial number (say 11), and a final number (say 14), the child is require to find the amount added - that is 11 + ? = 14. To put this in explicit form, it is required to state what operation must be carried out with the numbers 11 and 14 in order to generate the answer in the explicit form. Conventionally, this operation is again one of subtraction: i.e. 14 - 11 (the difference between 11 and 14). Note however, that the operation required (and found) is 'add 3'; whereas the meta-operation required to generate this (former) operation is conventionally known as subtraction. Note that transformational arithmetic distinguishes the operation which reverses the adding-on action, from the metaoperation used to find what that adding-on action is/was (Note 3).
Implications
The following are important reasons for studying counting signs:1. There is a need to signify the basic counting techniques of count-onlback and count-up; without them we cannot conveniently instruct children to adopt a certain strategy. Nor can children record which of these activities they carried out. This is clearly a disadvantage when attempting to develop a pedagogy for teaching counting skills. 2. Counting signs could be used to describe the counting strategies of non-literate cultural societies such as the the Kpelle of Liberia (Gay & Cole (1967).
l
2..\ Page 121
3. Since counting appears to be an intuitive strategy adopted by children prior to formal instruction, it would be of interest to psychologists and mathematicians to investigate the nature and properties of such signs.
Notes: 1. The 'counting signs' are currently being trialled in a National Numeracy Project school. 2. The counting procedures used by the Oksapmin involved what we could call a 'pre-counting-up' strategy in which the the names of the body parts or positions between two given positions were listed (orally) rather than counted. A more usual counting-up procedure is to number these 'counted between' body parts by calling them by other body parts. This is the standard procedure for counting-up between say 7 and II by enumerating 8,9, 10, 11 giving a count-up of 4.] In the model, counting-up from H to M is equivalent to finding how many letter-jumps between H and M. This is achieved by pointing to I, J, K, L, M and matching the pointing to the letters A, B, C, D, E . (I.e. since E is the 5 th letter, there are 5 letter-jumps between I and M.) 3. A fuller explanation of 'transformational arithmetic' is in preparation.
References Gay, J. & Cole, M. (1967) The new mathematics and an old culture. Holt Rhinehart & Winston, NY. Kaput, J.J. (1987) Symbol System Theory. In: C.Janvier (Ed) Problems of representation in the teaching and learning of mathematics. Hillsdale NJ. Laurence Erlbaum. Pratt, C. & Garton, A.F. (1993) Systems of Representation in Children: development and use. J. Wiley & Sons Ltd. Saxe, G.B. (1982) Culture and the development of numerical cognition: Studies among the Oksapmin of Papua New Guinea. In: c.J. Brainerd (Ed). Children's Logical and Mathematical Cognition. Springer-Verlag. NY. Sfard, A. & Linchevski, L. (1994) The Gain and Pitfalls of Reification - the case of Algebra. Educational Studies in Mathematics. 26. 191-228. Womack, D. (1998) Children inventing signs for their transformational number strategies. Proceedings of BSRLM conference in Oxford.
Page 122
Geometry Working Group

Convenor: Keith Jones, University of Southampton, UK
Visualisation, Imagery and the Development of Geometrical Reasoning

A report based on the meeting at the University of Birmingham, 20th June 1998 by Keith Jones, University of Southampton and Chris Bills, University of Warwick
This report focuses on some aspects of the nature and role of visualisation and imagery in the teaching and learning of mathematics, particularly as a component in the development of geometrical reasoning. Issues briefly addressed include the relationship between imagery and perception, imagery and memory, the nature of dynamic images, and the interaction between imagery and concept development. The report concludes with a series of questions that may provide a suitable programme for research and lays the foundation for further work of the BSRLM geometry working group.
The nature and role of visualisation and imagery in the teaching and learning of mathematics are complex. Much has been written about the value of visualisation and imagery in terms of the potential to enhance a global and intuitive view and understanding of various areas of mathematics (Bishop 1989, Fischbein 1987, Usiskin 1987, Zimmermann and Cunningham 1991). Fischbein (1987 p 1 04), for example, comments that "a visual image not only organises the data at hand in meaningful structures, but is also an important factor guiding the analytical development of a solution". Bishop (1989) concludes his review by saying that "there is value in emphasising visual representations in all aspects of the mathematics classroom" . Yet it is also recognised that there are difficulties concerned with visualisation and imagery (Dreyfus 1991, Love 1995). If mathematical visualisation is taken to be "the process of forming images (mentally, or with pencil and paper, or with the aid of technology) and using such images effectively for mathematical discovery and understanding" (Zimmermann and Cunningham 1991 p3), then such difficulties can relate to the process of forming images as well as using them in solving problems. Similarly, if mental imagery is taken as involving: "constructing an image from pictures, words or thoughts; re-presenting the image as needed; and transforming that image" (Wheatly 1991), then difficulties can arise from the processes of constructing, representing, and transforming. Love (1995 P 125) suggests that in geometry the relationship between "mental objects and physical images is an especially difficult one". From a slightly different perspective,
Page 123
Dreyfus (1991) comments on the low status often accorded to visual aspects of mathematics in the classroom. Recognising the complex nature of visualisation and imagery, especially its role in the development of geometrical reasoning, this paper presents a consideration of the visualisation process and the images formed. The French psychologist Raymond Duval (1998 p39) has suggested that: " differentiating between different visualisation processes ... is needed in the curriculum". So our central questions are: What are the different visualisation processes? What are the different types of mental images formed? We begin with an outline of the role of visualisation in the model of the development of geometrical reasoning proposed by Duval (1998 p38-39), which is under consideration by the BSRLM Geometry working group (Jones 1998). This leads to a consideration of various aspects of visualisation and imagery in mathematics education including the relationship between imagery and perception, imagery and memory, the nature of dynamic images, and the interaction between imagery and concept development. While this discussion raises more questions than it can answer, the resulting questions provide a suitable programme for research and lays the foundation for further work of the BSRLM geometry working group.
The Role of Visualisation in the Development of Geometrical Reasoning Duval suggests that geometrical reasoning involves three kinds of cognitive processes which fulfil specific epistemological functions. The three cognitive processes are : visualisation processes, for example the visual representation of a geometrical statement, or the heuristic exploration of a complex geometrical situation. construction processes (using tools) reasoning processes - particularly discursive processes for the extension of knowledge, for explanation, for proof Duval points out that these different processes can be performed separately. For example, visualisation does not necessarily. depend on construction. Even if a construction leads to a visualisation, construction processes, Duval contends, actually depend only on the connections between relevant mathematical properties and the constraints of the tools being used. Similarly, visualisation can be an aid to reasoning (for instance by aiding the finding of a proof) but visualisation can also be misleading (if our visualised image is a special case, for example). Duval argues, however, that, "these three kinds of cognitive processes are closely connected and their synergy is cognitively necessary for proficiency in geometry" (ibid p38). Duval illustrates the connections between these three kinds of cognitive processes as represented in Figure 1.
(lH- From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 124
In Figure 1 each arrow represents the way one kind of cognitive process can support another kind in any geometrical activity. Duval makes the arrow from visualisation to reasoning dotted because, as argued above, visualisation does not always help reasoning. The 'circular' arrow illustrates that reasoning can develop in a way that is independent of construction or visualisation processes. Given that the synergy of these three processes is cognitively necessary for proficiency in geometry, the issue, as identified by Duval, is how to get pupils in school to see the communication between the three kinds of processes. Duval argues that, in attempting to understand the development of geometrical reasoning, his research has shown the following: 1. The three kinds of processes must be developed separately. 2. Work on differentiating visualisation processes and between different reasoning processes is needed in the curriculum. 3. The co-ordination of these three kinds of processes can really occur only after this work on differentiation.
Visualisation and Imagery Processes in Mathematics This section attempts to summarise some of the issues concerned with visualisation, which may be mental or physical, and imagery, which may be pictorial. In particular we consider the relationship between imagery and perception, imagery and memory, the nature of dynamic images, and the interaction between imagery and concept development. imagery and perception While Duval may appear to prioritise visual perception (his area of expertise in psychology), 'seeing' is not the only source of mental imagery in mathematics. 'Feeling' physical objects without looking (in other words, touch perception) is another source of mental image creation. An exploratory study involving 3D geometrical objects undertaken by Triadafillidis (1995) shows some of the potential. It is worth noting,
however, that there is no consensus about how perceptions are coded by

the mind, nor how these codes are represented mentally. Love (1995 p 125)
\'lS Page 125
suggests that there is disagreement "over whether such things as 'pictures in the mind' can exist independently of thought and language or even whether they exist at all". A question that we do not have space to address here is whether visualisation even needs sight. imagery and memory Another open question is the relationship between memory and imagery, or perhaps better, the role of memory in imagery and visualisation. Are mental images formed from visual experiences necessarily pictures that can be viewed in the mind, or simply memories of that experience? Presmeg (1986), in a study of what she called 'visualisers' (those who prefer to use visual methods when attempting to solve mathematical problems that could be solved by both visual and non-visual methods), identified five kinds of visual imagery, which she referred to as: pictorial (picture-in-the-mind) pattern (relationships depicted spatially) memory (recreating images from experience) kinesthetic (involving muscular activity) dynamic (moving) In Presmeg's classification an image recreated from memory of a visual experience mayor may not be pictorial. dynamic images As Presmeg identified, some visual thinkers are able to make use of dynamic mental images. The impact of the forms of dynamic diagrams available in computer-based mathematical learning environments on the development of such imagery is not known, although Gorgori6 and Jones (1996) suggest that the use of a dynamic geometry package such as Cabrigeometre can support the development of important visualisation skills necessary for the understanding of visual phenomena. imagery and concept development While many have suggested that the use of imagery aids conceptual development, there is still some way to go to understanding the precise relationship. Mariotti (1995 pi 04) suggests that geometrical reasoning can be interpreted in terms of "a dialectical process between the figural and conceptual aspects". In other words, geometrical reasoning involves an inter-dependent relationship between images and concepts. Simpson and Tall (1998) make the distinction between passive, organisational, and generative figures, we could use the same classification for mental images. A passive image could be merely associated with a concept whilst an organisational image allows information to be represented compactly. Alternatively a generative image is used by the learner to guide their learning and it may be conceptually or formally generative. In geometry, the passive image of a regular pentagon that many learners visualise can positively inhibit the development of the concept of a
\2Jo From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author
Page 126
pentagon as any five sided shape. Images of objects being "dragged" as in a dynamic geometry package, on the other hand, can enhance that conceptual development and could thus be conceptually generative. Formally generative imagery would include: a "proofwithout-words" of Pythagoras' Theorem, the visualisation we might employ to guide a formal proof or a visualised "sketch" used to generate a more formal construction.
Conclusions While the above discussion probably raises more questions than it can answer, the resulting questions provide a suitable programme for research and the foundation for further work of the BSRLM geometry working group. To paraphrase Drefyfus (1995 P 16-17), the overarching need is for theory building, with input both from classroom experiences and carefullydesigned research. In particular we need: to understand the precise role of diagrams in problem solving and learning about specific mathematical concepts and processes to find out for what kinds of reasoning processes and in what kinds of learning situations, diagrams and/or visual imagery are particularly helpful to understand the impact on mathematical reasoning of dynamic diagrams available in computer-based mathematical learning en vironmen ts to find out what are efficient means for communication about, and by means of, diagrams, and their associated interpretations.
References Bishop, A J (1989), Review of Research on Visualisation in Mathematics Education. Focus on Learning Problems in Mathematics. 11(1), 7-16. Dreyfus, T (1991), On the Status of Visual Reasoning in Mathematics an~ Mathematics Education. In: F Furinghetti (Ed), Proceedings of the 1St Annual Conference of the International Group for the Psychology of Mathematics Education. Assissi. Vol 1, 3348. Dreyfus, T (1995), Imagery for Diagrams. In R Sutherland and J Mason (Eds), Exploiting Mental Imagery with Computers in Mathematics Education. Berlin: Springer. Duval, R (1998), Geometry from a Cognitive Point of View. In C Mammana and V Villani (Eds), Perspectives on the Teaching of Geometry for the st 21 Century: an ICMI study. Dordrecht: Kluwer. Fischbein, E (1987), Intuition in Science and Mathematics. Dordrecht: Kluwer. Gorgori6, N and Jones, K (1996), Elements of the Visualisation Process within a Dynamic Geometry Environment. Invited paper presented to the topic group on The Future of Geometry at the 8th International Congress on Mathematical Education, Seville, Spain, July 1996. 6pp.
12. 7
Hershkowitz, R. (1990). Psychological Aspects of Learning Geometry. In P. Nesher & J. Kilpatrick (Eds.), Mathematics and Cognition. (pp. 70-95). Cambridge: CUP. Jones, K (1998), Theoretical Frameworks for the Learning of Geometrical Reasoning. Proceeding of the British Society for Research into Learning Mathematics. King's College, University of London, February 1998. Love, E (1995), The Functions of Visualisation in Learning Geometry. In R Sutherland and J Mason (Eds), Exploiting Mental Imagery with Computers in Mathematics Education. Berlin: Springer. Mariotti, M A (1995), Images and Concepts in Geometrical Reasoning. In R Sutherland and J Mason (Eds), Exploiting Mental Imagery with Computers in Mathematics Education. Berlin: Springer. Presmeg, N C (1986), Visualisation in High School. For the Learning of Mathematics. 6(3), 42-46. Simpson, A and Tall, D (1998), Computers and the Link Between Intuition and Formalism. In print Triadafillidis, T A (1995), Circumventing Visual Limitations in Teaching the Geometry of Shapes. Educational Studies in Mathematics. 29(3), 225-235. Usiskin, Z (1987), Resolving the Continuing Dilemma in School Geometry. In: M M Lindquist and A P Shulte (Eds), Learning and Teaching Geometry K12. Reston, VA: NCTM. Wheatly, G (1991), Enhancing Mathematics Learning through Imagery. Arithmetic Teacher. 39(1), 34-36. Zimmermann, Wand Cunningham, S (1991), Visualisation in Teaching and Learning Mathematics. Washing: Mathematical Association of America.
BSRLM Geometry Working Group The geometry working group focuses on the teaching and learning of geometrical ideas in its widest sense. The aim of the group is to share perspectives on a range of research questions which could become the basis for further collaborative work. Suggestions of topics for discussion are always welcome. The group is open to all.
Contact: Keith Jones, University of Southampton, Research and Graduate School of Education, Highfield, Southampton, S017 1BJ, UK. e-mail: dkj@southampton.ac.uk tel: +44 (0) 1703 592449 fax: +44 (0) 1703 593556 http://www.soton.ac. uk/- gary/crime.html
Page 128
Semiotics and Mathematics Education Working Group

Adam Vile, South Bank University and Paul Emerst, Exeter University
The theme of this meeting was "applied semiotics" and its role in mathematics education research. There were presentations from two researchers working in the context of semiotic methodologies and perspectives specifically aimed at discussion of the tools and techniques that they were using for their data collection and analysis. Dedrie Cook (Derby) spoke about her work in studying children at play with mathematical artefacts and the tools from linguistics that she had used for analysis of data. Corina Silveira (Southampton) shared her theoretical perspective and research design for analysis of children's development of counting. The work of Dedrie and Corina exemplified the great variety in work that is considered semiotic, and presented two approached to data collection and design that fit within a semiotic framework. In the following discussion period the group was asked to consider the following questions, none of which were resolved (of course this was not the intention): ~ Does it make sense to talk about applied semiotics? ~ What toolkits have been used? ~ What toolkits are available? ~ In what sense are existing tools examples of applied semiotics? ~ Can these tools add anything new to our understanding of mathematics teaching and learning? ~ Why adopt a semiotic approach at all? ~ How do we view validity and reliability in this context? The future of the group was discussed and there was a clear consensus that the group should continue, perhaps in a more focussed way. More use will be made of the mailing list and it was suggested that readings were posted prior to each meeting of the group that will focus the discussion. It was also suggested that the group begin, slowly, a writing project that will bring semiotic ideas to the mathematics teaching community in a practical and accessible way. Perhaps through Mathematics teaching publications. The reading for the next meeting will be decided through the semimath mailing list and published one month prior to the meeting. To subscribe to the mailing list send the following message to maiordomo@sbu.ac.uk: subscribe semimath youremail@wherever.
Page 129
Page 130

Duval.29. Razonamiento Geometrico

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Duval.29. Razonamiento Geometrico

Hochgeladen von

Copyright:

Verfügbare Formate

British Society for Research into Learning Mathematics

Working Group Reports from February 28th 1998

Primary Children's Imagery in Arithmetic Sandra Pendlington University of Exeter

Research Reports presented on June 20th 1998

Working Group Reports from June 20th 1998

Mathematics in the Chemistry course at Imperial College

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Backgrounds of the students

A comparison of two students

little in school, but computers not at all.

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Issue 1: Confidence and facility

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Issue 2: Explicit and implicit knowledge, "intuition"

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Issue 3: "What do you think you've learnt?"

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

3. Current use of Investigatory approaches in Teacher training in India

BSRLM 28/ 02/98

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Findings from Questionnaires

8. Implications for teacher training in India

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Geometry Working Group

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

level 2 level 3 level 4

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Part I: What has AMT offered to Mathematics Education?

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

1. Concept Image/Concept Definition. 2.Reflective Abstraction. 3. Procept.

Part II: How can we re-shape AMT?

to work towards a clearer understanding of the nature of advanced mathematical thinking.

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

Mapping from representation to image

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author

From Informal Proceedings 18-1&2 (BSRLM) available at bsrlm.org.uk the author