British Educational Research Journal

Vol. 32, No. 1, February 2006, pp. 3955

The effect of grammar teaching on

writing development
Richard Andrews*a, Carole Torgersona, Sue Bevertonb,
Allison Freemana, Terry Lockec, Graham Lowa,
Alison Robinsona and Die Zhua
a

University of York, UK; bUniversity of Durham, UK; cWaikato University, New

Zealand
(Submitted 2 November 2004; resubmitted 24 March 2005; accepted 27 May 2005)

This article reports on the results of two international systematic research reviews which focus on
different aspects of teaching grammar to improve the quality and accuracy of 516-year-olds
writing in English. The results show that there is little evidence to indicate that the teaching of
formal grammar is effective; and that teaching sentence-combining has a more positive effect. In
both cases, however, despite over a hundred years of research and debate on the topic, there is
insufficient quality of research to prove the case with either approach. More research is needed, as
well as a review of policy and practice in England with regard to the teaching of sentence structure
in writing.

Introduction
The question of whether the teaching of grammar has a positive effect on young
peoples writing development has been haunting policy-makers, teachers, researchers and students themselves for over a hundred years. The haunting is real for
students in England who have been on the receiving end of a form of teaching that
previous reviews and primary studies have suggested is ineffective. It is interesting
that a belief in teaching formal grammar has persisted in the face of the evidence:
perhaps there is a (mistaken) belief that a description of a linguistic system (a
sentence grammar) can be translated into textbooks and pedagogies that, it is hoped,
will benefit young peoples writing development. Although there have been extensive
reviews of the question (e.g. Macaulay, 1947; Wilkinson, 1971; Wyse, 2001), views
remain polarized, with a belief among some teachers, newspapers and members of
the public that such teaching is effective, and among others that it is ineffective.
*Corresponding author. Department of Educational Studies, University of York, York YO10
5DD, UK. Email via Alison Robinson: ar31@york.ac.uk
ISSN 0141-1926 (print)/ISSN 1469-3518 (online)/06/010039-17
# 2006 British Educational Research Association
DOI: 10.1080/01411920500401997

40 R. Andrews et al.
In an attempt to shed some conclusive light on the matter, the English Review
Group at the University of York, in association with the Evidence for Policy and
Practice Information and Coordinating Centre (EPPI-Centre), undertook a
systematic review in 2004 to answer the research question, What is the effect of
grammar teaching in English on 516-year-olds accuracy and quality in written
composition? It did so with two objectives in mind: to map the field of research on
the effects of grammar teaching on writing in English-speaking countries for pupils
aged between 5 and 16; and to undertake an in-depth review of two aspects of the
field (the teaching of formal sentence grammar/syntax, and the teaching of a
technique known in the USA as sentence-combining). This article presents the
results of the two in-depth reviews.
Background
Since the publication of the Kingman Report (Department of Education and
Science, 1988) there has been a conviction amongst curriculum writers and
policy-makers in England that grammar teaching to young learners of English is
a good thing; that it will improve their written English and their ability to talk
about language; that talking about language is helpful in understanding language and, in turn, in improving its use; and that such reflection and discussion
about language should start earlier than had previously been thought possible or
desirable.
It should be said at the start that such a conviction flies in the face of much
research evidence from the twentieth century (partly summarized in Wilkinson
[1971, pp. 3235]). Perera (1984) noted that decontextualized grammar teaching
that was unrelated to pupils other language work was likely to do more harm than
good. She also noted that technical terms in grammar seemed to confuse rather than
enlighten young people.
Wilkinson notes that although grammar is a useful descriptive and analytical tool,
other claims made for it are nearly all without foundation (1971, p. 32). Studies in
the twentieth century have suggested that the learning of formal, traditional (i.e. not
transformative) grammar has no beneficial effect on childrens written work (Rice,
1903); that training in formal grammar does not improve pupils composition
(Asker, 1923; Macaulay, 1947; Robinson, 1960); that ability in grammar is more
related to composition in some other subjects than in English (Boraas, 1917; Segal &
Barr, 1926); that a knowledge of grammar is of no general help in correcting faulty
usage (Catherwood, 1932; Benfer, 1935); that grammar is often taught to children
who have not the maturity or intelligence to understand it (Symonds, 1931;
Macaulay, 1947); and that teaching grammar may actually hinder the development
of childrens English (Macaulay, 1947).
A recent critical review of the empirical evidence on the teaching of grammar
provided an overview of research studies in English-speaking countries (Wyse, 2001).
This review concluded that the teaching of grammar (using a range of models) has
negligible positive effects on improving secondary pupils writing (p. 422).

Effect of grammar teaching on writing development 41

Policy and practice in the 1970s and 1980s in England followed a line
characterized by the Bullock Report (Department of Education and Science,
1975), that it was teachers who needed to know about grammatical construction
so that they could understand pupils writing problems and intervene accordingly
and appropriately, with language competence growing incrementally through
the interaction of reading, writing, speaking and listening. More recently in
England and Wales, the National Literacy Strategy (which operated for 711year-olds from 1997 before being extended to 1114-year-olds in 2002) has issued
a book and video entitled Grammar for writing (Department for Education
and Employment, 2000), aimed particularly at the teaching of 711-year-olds.
The basic principle behind this relatively recent initiative is that all pupils
have extensive grammatical knowledge (p. 7) and that teaching that focuses
on grammar helps to make this knowledge explicit. Such explicitness, so the
book and video argue, helps to improve young peoples writing through providing
them with an increase in the range of choices open to them when they write (p. 7).
Throughout, there is a distinction between spoken grammars and written grammars, and a clear objective to support the development of a command of
sentence construction. In pedagogical terms, the emphasis of the book is on
teaching at the point of compositionwhat Britton (1983) called shaping at the
point of utterance.
While eschewing a return to the descriptive and prescriptive grammar teaching of
the 1950s and 1960s, the current orthodoxy focuses clearly on the improvement of
sentence structure and uses extensive knowledge about language and increased
language awareness as a means to help pupils to write better English. It consists of a
detailed programme for using sentence grammar to improve sentence construction,
via explicit teaching. As such, it represents a middle ground between traditional
grammar teaching on the one hand, and language awareness arising from the use of
language in speech and writing on the other. This approach is based on two
publications from the Qualifications and Curriculum Authority (QCA) which
enshrine recent government thinking in England and Wales, and which underpin
teaching approaches in the National Curriculum: The grammar papers (QCA, 1998)
and Not whether but how (QCA, 1999), the latter title revealing a worrying conviction
that it is not worth asking the question as to whether grammar should be taught in
the curriculum. It is interesting to note that in response to press releases about the
two grammar reviews we discuss in the present article, the Department for
Education and Skills (DfES) took the line of eclecticism: a wide range of methods
and approaches was encouraged. Our concern is that not only is the implementation
of the National Literacy Strategy patchy, but the very wide range that is extolled as a
virtue by the DfES is actually leading to confusion about the best way forward in
improving young peoples writing.
It is fully acknowledged in the present reviews that sentence-level grammar is
contingent upon the levels of text grammar (above the level of the sentence) and
word grammar (below the level of the sentence). Our aim was to focus on sentencelevel operations in teaching about writing and in learning to write.

42 R. Andrews et al.
Definitions and some caveats
In any research of this scope, it is important to define the terms used, both for
claritys sake but also to define the parameters within which the research review
operated.
Grammar, as far as the present review was concerned, refers to written sentence
grammar. It includes the study of syntax (word order), clause and phrase structure,
and the classification of parts of speech (e.g. noun, verb, predicate, clause, etc.) It
can be both descriptive, in that it describes the existing patterns of sentences; and
also generative or transformative, in that rules can be defined which can generate
grammatically acceptable sentences (the transformation being from basic rules
through to actual sentences). Studies of words or subcomponents of words are not
part of the study of grammar per se. Similarly, studies in language awareness are not,
strictly speaking, part of the present review, though we discuss its nature and
function during the course of the article. We also concede that this definition of
grammar is narrower that the one used by Cope and Kalantzis (1993) where
grammar is a term that describes the relation of language to metalanguage; of text
to generalizations about text; of experience to theory; of the concrete world of
human discursive activity to abstractions which generalize about the regularities and
irregularities of that world (p. 20).
By written composition, we meant extended pieces of writing (in handwriting, in
type or via word processing) in a variety of genres or text-types. In focusing on
accuracy we meant to place emphasis on appropriateness of grammatical form for
particular purposes. We were not concerned with spelling accuracy, nor with
legibility, neatness of handwriting or vocabulary (except where it bore upon sentence
grammar). The emphasis on quality (e.g. syntactic maturity) was there to distinguish
our study from an interest in quantity (e.g. the number of sentences or components
of sentences in a composition).
By English-speaking countries we meant countries where English is spoken and
written as a first language by a significant segment of the population. Included were
the UK, Ireland, USA, Canada, Australia, New Zealand, Jamaica and other
countries in the Caribbean, Gibraltar and South Africa.
We defined syntax as: constraints which control acceptable word order within a
sentence, or dominance relations (like head noun+relative clause); and sentencecombining as meaning: teaching techniques for splicing together simple sentences to
make compound or complex ones. It can also cover sentence-embedding and other
techniques for expanding and complicating the structure of sentences.
Methods
This systematic review used guidelines and tools devised by the EPPI-Centre (EPPICentre, 2002a, b, c). The protocol (published on the Research Evidence in
Education website: http://eppi.ioe.ac.uk/reel) included the review questions and

Effect of grammar teaching on writing development 43

inclusion/exclusion criteria for the searching and screening of potentially relevant
studies. Studies were included in the systematic map if they looked at the effect of
grammar teaching in English on 516-year-olds accuracy and quality in written
composition. As the focus of the study is on the effects of grammar teaching, papers
using methods to identify any such effects were required.
The studies remaining after application of the criteria were keyworded using the
EPPI-Centres Core Keywording Strategy (EPPI-Centre, 2002a) and online database
software, EPPI-Reviewer (EPPI-Centre, 2002b). Additional review-specific keywords which are specific to the context of the review were added to those of the
EPPI-Centre. Again, quality assurance (QA) was provided by the EPPI-Centre.
A map of research studies in the field was generated. From the map, two areas of
research were identified for in-depth review: formal grammar (syntax) and sentencecombining.
Studies identified as meeting the inclusion criteria for the in-depth review were
analysed in depth using the EPPI-Centres detailed Data-Extraction Guidelines
(EPPI-Centre, 2002c), together with its online software, EPPI-Reviewer H (EPPICentre, 2002b). All studies were double data extracted by two reviewers working
independently, who then discussed any differences and resolved them. Three
components were identified to help in making explicit the process of apportioning
different weights to the findings and conclusions of different studies. Such weights of
evidence are based on the following:
(a) the soundness of studies (internal methodological coherence), based upon the
study only;
(b) the appropriateness of the research design and analysis used for answering the
review question;
(c) the relevance of the study topic focus (from the sample, measures, scenario, or
other indicator of the focus of the study) to the review question;
(d) an overall weight taking into account (a), (b) and (c).
The data were then synthesized to bring together the studies which answer the
review question and which meet the quality criteria relating to appropriateness and
methodology. A narrative synthesis was undertaken.
Criticisms have been levelled at the above approach from a number of
perspectives, though many of these have been about the nature of evidence and its
relationship with policy-making. Thomas and Pring (2004), for example, collect a
range of essays for and against systematic reviewing of research evidence in their
excellent Evidence-based practice in education, which is discussed in depth in Andrews
(2005) in relation to the place of systematic reviews in educational research. With
regard to the present reviews, one could mount a number of specific criticisms: that
the studies did not take into account the cultural, social and pedagogical issues
associated with grammar teaching; that qualitative data was not sufficiently
included; that research carried out in the USA or New Zealand may not apply to
the current British context; that the focus of the reviews is too narrow.
We will answer these criticisms briefly here, as a full-scale debate is conducted in
Thomas and Pring (2004) and would take at least another article, if not book. Part of

44 R. Andrews et al.
the approach of systematic reviews is that they work from a finely honed research
question. Our own main research questionWhat is the effect of grammar teaching
in English on 516-year-olds accuracy and quality in written composition?was
generated by a steering group consisting of teachers, policy-makers, governors,
parents, teacher-educators and researchers. It then went through a process of
refinement. The question is a specific one, looking for effect and effectiveness. It
therefore required a particular kind of methodology, taking into account cultural,
social and other contextual issues as they pertained to each of the experimental
studies we reviewed. We were looking to gauge the effectiveness of a particular
pedagogical approach and to generalize the results through synthesis, not to give
finely grained accounts of the teaching of grammar in different contexts. Because the
phenomena we were investigatingthe teaching of formal grammar and sentencecombiningcould be isolated from context (that perhaps is part of their problem!)
the methodological approach was appropriate. One could argue that we have
narrowed down the field of potential insights, and we would accept such a criticism;
at the same time, we hope to have shed more conclusive light than before on two
particular aspects of the teaching of first-language writing in English to pupils aged
516 across the world.
A further criticism could be made that the process of distilling from an initial large
number of found studies in the field to a relatively small number examined in depth
is over-selective. Such a criticism can be countered by the fact that the process is
transparent and explicit: all excluded studies are listed in the full technical reports,
and each stage of distillation is not only undertaken by more than one researcher, but
also carefully recorded. The process of including or excluding studies should be
replicable.
Results
The initial electronic searching for research in the field between 1900 and 2004
identified 4691 papers, which were screened for potential relevance on the basis of
title and abstract. A further 50 potentially relevant papers were identified through
hand searching. A total of 267 papers was obtained and all were re-screened on the
basis of the full paper. Of these, 64 turned out to be relevant to the particular criteria
for the review and constituted a map of the field. Twenty-six papers reported reviews
and 38 reported primary research. Of the latter group, 11 papers were deemed by the
review group to be highly relevant to the in-depth review on syntax and 20 papers,
reporting on 18 studies, were relevant to the in-depth review on sentencecombining.
Syntax
Eleven studies were identified for the in-depth review on syntax. The 11 studies
selected for in-depth review were all experimental studies, of which three were
randomized controlled trials (Bateman & Zidonis, 1966; Thompson & Middleton,

Effect of grammar teaching on writing development 45

1973; Fogel & Ehri, 2000); two were controlled trials (Elley et al., 1975, 1979;
Stock, 1980); four used pre- and post-tests (Hilfman, 1970; Rousseau & Poulson,
1985; Roberts & Boggase, 1992; McNeill, 1994); one was a curriculum evaluation
(Satterfield & Powers, 1996); and one a single subject ABACA design (Stone &
Serwatka, 1982).
The narrative overview in the full report, and in the present article, focuses on the
studies rated high and high/medium or medium/high in quality. These are Elley et al.
(1975, 1979) (high to medium); Bateman and Zidonis (1966) (medium to high);
and Fogel and Ehri (2000) (high).
The two Elley et al. papers are based on a single study, the aim of which was to
carry out a study of the effects of traditional and transformational grammar on
childrens writing skills, and in so doing to avoid the deficiencies of previous research
on the subject (Elley et al., 1975, p. 27). The investigation, a controlled trial, was
conducted in one co-educational high school in Auckland in the early 1970s. Within
the school, 248 pupils in eight matched classes of average ability were taught,
observed and regularly assessed from the beginning of their third-form year to the
latter part of their fifth-form year (Elley et al., 1975, p. 28), i.e. from age 13/14 to
15/16. Elley et al. describe how, at the outset, one bright and three slow-learning
classes were deliberately excluded from the study, thus rendering [the sample] more
homogenous, and increasing the chance of identifying systematic differences
between groups (p. 28). The experimental pupils were classified into eight matched
classes of 31 pupils on the basis of a number of tests, and additional matching
criteria were ethnic group, sex, contributing school, and subject options (p. 28).
Although the pupils were allocated as individuals to the eight classes, the study
after this allocationworks as a cluster trial as the pupils in the eight classes were
divided into 3, 3 and 2 classes, and the pupils tested during the intervention period
and at the end.
The three courses studied by the three groups were, essentially, a transformational
grammar course; a reading-writing course, which substituted extra reading and
creative writing for the transformational grammar strand (Elley et al., 1975, p. 29);
and a traditional grammar course. Each cluster of classes was taught one of these
methods. In summary, the results of the study reveal that in Form 3 (nowadays Year
9, i.e.13/14-year-olds), on none of the 12 variables chosen for analysis did any
English programme show a significant superiority. The transformational grammar
classes liked writing less than the other groups. In Form 4 (Year 10, 14/15-yearolds), only one comparison (from 30 possible) showed significant differences (on
essay content). In the School Certificate Examination, there were no significant
differences between the three programmes. In Form 5 (Year 11, 15/16-year-olds),
only two of the 12 variables listed showed any significant differences (sentencecombining test and English usage test). Again, in the School Certificate
Examination, there were no significant differences between the three groups. In
attitude tests, transformational grammar pupils found English lessons more
repetitive and useless than the other groups. Elley et al. conclude that English
grammar, whether traditional or transformational, has virtually no influence on the

46 R. Andrews et al.
language growth of typical secondary school students (1975, p. 38). Although they
suggest that the benefits to be derived from grammar teaching might be confined to
bright children, i.e. those who can readily manipulate and apply those syntactic
structures which they studied (p. 39), the study had deliberately omitted the top 8
10% of pupils in the year and the suggestion remains a speculation. Nor can the
results be generalized to older students, or to learners of English as a second
language.
The other studies rated relatively highly by the review team in relation to the
specific focus of our study were those by Bateman and Zidonis (1966) and Fogel and
Ehri (2000). Bateman and Zidonis examined a cohort of pupils moving from ninth
to tenth grade in an American high school. Fifty pupils were assigned randomly to
two sections and teachers were assigned randomly to the two classes. Although this
is an individually randomized controlled trial, in that pupils were assigned randomly
to the two groups, thereafter the study works as a cluster trial because the pupils in
each class were taught together. The study sought to measure the effect that the
teaching of a generative grammar had upon the writing of pupils, aiming also to help
them become stylists who have expanded their capability of generating varied and
well-formed sentences of the language (Bateman & Zidonis, 1966, p. ix). The
experimental group was required to learn from special grammatical materials
provided by the investigators. Written compositions were collected from both groups
during the first three months of the first year and the last three months of the second
year of the project. Results show that there was a greater increase in the average
structural complexity scores for well-formed sentences in the experimental group
than in the control group; and that measures of five grammatical operations
indicated that the experimental group had the edge. However, the greatest changes
were made by only four students, one of whom showed a very large increase in
structural complexity. Similarly, the experimental group was better able to hold in
check an increase in malformed sentences as structural complexity increased.
In terms of the research question we are trying to answer, a relatively welldesigned randomized controlled trial like this would appear to give us the most valid
and reliable results. However, Bateman and Zidonis conclude, quite rightly, that the
findings should be treated with caution because, although criteria of internal validity
were met through careful randomization procedures, the sampling requirements
needed to meet criteria of external validity could not be adequately fulfilled.
Furthermore, the analyses do not take the clustered nature of the data into account.
Nevertheless, the study suggests, tentatively, that high school students can learn the
principles of generative grammar; that a knowledge of generative grammar enables
such students to increase the proportion of well-formed sentences they write; and
that a knowledge of generative grammar can enable students to reduce the
occurrence of errors in writing.
Fogel and Ehris (2000) aim was to examine how to structure dialect instruction
so that it is effective in teaching Standard English forms to students who use Black
English Vernacular in their writing (p. 215). This is thus a very different study from
the previous two, discussed above, in that it does not look at transformational/

Effect of grammar teaching on writing development 47

generative or traditional grammar teaching. Rather, its focus is on third and fourth
grade (8/9-year-old) African-American pupils who exhibited Black English
Vernacular (BEV) features in their written work and yet who must acquire writing
competence in Standard English (SE). Three training procedures were designed to
teach elementary school pupils six SE syntactic forms: the Exposure-only (E)
condition, in which pupils listened to stories that included multiple instances of the
SE forms; the Exposure/Strategies condition (ES), in which pupils listened and in
which the teacher labelled and illustrated use of the six SE forms; and the Exposure/
Strategies/Practice (ESP) condition, in which pupils added to the above procedures
by translating sentences from BEV to SE and then received feedback from the
teacher on their accuracy. As Fogel and Ehri describe, effects of the treatments were
assessed with measures of students ability to translate BEV sentences into SE
sentences and their ability to use SE forms in their free writing. Students confidence
in their ability to write using correct grammatical or SE forms was assessed with a
self-efficacy measure (Fogel & Ehri, 2000, pp. 227228).
The results of this study showed that the ESP treatment was more effective in
teaching students to write using SE forms than either of the other two treatments.
What the study reveals is, perhaps, more pedagogical than grammar-based:
differences between BEV and SE are grammatical issues, but it is not until such
differences are understood and then practised in writing that they take effect.
It is not possible to synthesize systematically the results of the studies by Elley et al.
and Bateman and Zidonis. First, the transformational grammatical approach of Elley
et al., based as it is on materials from the Oregon Curriculum (Kitzhaber, 1968),
useswe assumedifferent intervention materials from the unspecified special
grammatical materials of Bateman and Zidonis. Second, the analytical framework of
the two studies is different, with Elley et al. using 12 variables for analysis and
Bateman and Zidonis, 46. Third, we cannot rule out from either study, for different
reasons, methodological invalidity or unreliability. Fourth, there is insufficient detail
given in Bateman and Zidonis of the intervention or of the analytical tools used
(hence the lower rating than Elley et al. in terms of weight of evidence). Fifth, there is
no clear comparability between the two studies because Elley et al. use what they call
a transformational approach, and Bateman and Zidonis use a generative approach
to transformational/generative grammar. The relationship between the two, and to
transformational and generative grammars and theories, is not clearly articulated.
In summary, Elley et al. conclude that syntax teaching, whether traditional or
transformational, has virtually no influence on the language growth of typical
secondary school students. Bateman and Zidonis conclude, tentatively, that a
generative grammar approach does make a difference to syntactic quality and to the
control of malformed sentences. Because of the relative quality of the two studies,
methodologically, the results of the Elley et al. study have a higher weight of
evidence. However, neither study can be said to be conclusive. Fogel and Ehri
present a different kind of study in which mastery of standard English written forms
is improved for elementary school African-American pupils by a process of exposure,
strategies for labelling and identifying grammatical features and, crucially, practising

48 R. Andrews et al.
writing in these forms and receiving teacher feedback. However, short-term feedback
is not enough to cause change in pupils of this age.
Sentence-combining
In the first part of this article, we gave a simple definition of sentence-combining.
Essentially, it is a range of practical techniques for moving from existing sentences
and elements of sentences to compound and complex sentences. For example,
two simple sentences like The government is seeking more research evidence
and It thinks that such evidence will inform its policy-making can be combined
in a number of ways: with a connective (The government is seeking more
research evidence as/because it thinks that such evidence ), a semi-colon (The
government is seeking more research evidence; it thinks that such evidence ) or
in other ways (e.g. subordination: Because the government thinks research evidence
will inform its policy-making, it is seeking more such evidence.) As a technique,
it includes embedding (The British government is seeking more and better research
evidence as it thinks such evidence, if appropriately transformed will inform
its previously hit-and-miss, short-termist policy-making.) The embedding and
sentence-combining processes can work in reverse, by simplifying complex,
ill-expressed or ill-structured sentences. The main point that distinguishes
sentence-combining and its associated techniques from traditional formal
grammar teaching is that the former is practical: a technique used in specific
situations. The latter is abstracted from practice and usage, formulated into rules,
and then applied. We are not saying, by the way, that teachers of writing may not
need to know about formal grammar; they may, indeed, need to draw on such
knowledge in order to help their pupils to make appropriate choices in the act of
composing.
For the purposes of the narrative synthesis in the research review of sentencecombining, we took only those studies ranked of medium weight of evidence overall
or above. This does not mean to say that those studies ranked medium to low and
below are not worthy studies; it simply means that for the purposes of answering the
specific research questions in the present systematic review, those ranked medium
and above provide the best evidence.
OHares (1973) study arguably represents the best study so far on the effect of
sentence-combining on written composition. Its aim was to test whether sentencecombining practice that was in no way dependent on the students formal knowledge
of transformational grammar would increase the normal rate of growth of syntactic
maturity in the students free writing in an experiment at the seventh grade level over
a period of eight months (p. 35). Within a total sample of 83, students were
randomly assigned to two experimental and two control classes, thus creating a
randomized controlled trial. Students in the experimental groups were exposed to a
range of sentence-combining techniques of the kind set out above. Pre- and posttests were undertaken on three kinds of writing sample: narration, description and
exposition; and six factors of syntactic maturity were employed: words per T-unit,

Effect of grammar teaching on writing development 49

clauses per T-unit, words per clause, noun clauses per 100 T-units, adverb clauses
per 100 T-units and adjective clauses per 100 T-units. (T-units were defined by
Hunt [1965] as the shortest grammatically allowable sentences into which [writing
can be segmented]. To Hunt, a T-unit is essentially composed of a main clause
along with one or more subordinate clauses that go with it.) This particular study is
comprehensive, with high degrees of validity and reliability.
Results from the study show that not only did the experimental group experience
highly significant growth, but that its performance exceeded that of the control
group on all six measures of syntactic maturity. Indeed, eighth graders were writing
at the same syntactic maturity level as twelfth graders on five of the six measures. In
terms of writing quality, as judged and agreed by a team of eight evaluators, the
experimental group also exceeded the performance of the control group, particularly
in narrative and descriptive composition. The author of the study concludes that
teachers of writing surely ought to spend more time teaching students to be better
manipulators of syntax. Intensive experience with sentence-combining should help
enlarge a young writers repertoire of syntactic alternatives and to supply him [sic]
with practical options during the writing process (p. 76).
Saddler and Grahams study (2005) suggests that sentence-combining is not a
practice or a topic for research confined to the 1960s to 1980s. The aim of their
study was to examine the effectiveness of sentence-combining instruction, coupled
with peer instruction, for improving a basic foundation writing skill, sentence
construction (p. 4). The assumption behind this particular study is that facility in
generating sentences should make available more cognitive resources for other
aspects of composition. Using a sample of 44 911-year-old students (the mean
participating age was 9 years, 3 months), the authors used sentence-combining or
grammar interventions with pupils in pairs in laboratory-like conditions. The study
type was an individualized randomized controlled trial with stratified randomization;
baseline equivalence was used to eliminate chance bias. Although the reliability of
the study was high, validity would seem to be less strong than in OHare or other
studies. However, the results are clear: sentence-combining instruction was effective
in improving the sentence-combining skills (p. 29) and has a positive impact on
writing quality, not only in first versions of writing but also in subsequent revisions.
The effect of sentence-combining was seen to be stronger in the development of
syntactic maturity than in the improvement in writing quality. The writers conclude
that findings from the current study replicate and extend previous research by
showing that a peer-assisted, sentence-combining treatment can improve the
sentence construction skills of more and less skilled young writers and that such
instruction can promote young students use of sentence-combining skills as they
revise (p. 37).
OHare (1973) and Saddler and Graham (2005) were thought to have the highest
weight of evidence in relation to the research question set by the review. The
following studies provided medium weight of evidence overall. Of these, the studies
by Combs (1976, 1977) and Hunt and ODonnell (1970) were afforded medium to
high weight of evidence.

50 R. Andrews et al.
Combss studies replicated aspects of earlier studies by Mellon (1969) and
OHare (1973) with a seventh-grade sample of 100 students. The design of the study
included two intact experimental classrooms and two intact control classrooms
selected from a suburban Minneapolis junior high school and followed the pre-test
control group design excepting the random selection of the student population
and the inclusion of a delayed post-test. In effect, this was a clustered controlled
trial. Narrative and descriptive modes of writing were used to provide writing
samples and seven teacher-raters were used to gauge the quality of matched pairs of
writing from the control and experimental groups. The study was relatively well
conducted in terms of validity and reliability, and its results show that using words
per T-unit and words per clausethe two most discriminating measures in terms of
syntactic maturityrevealed that students made a grade leap of +2, as opposed to
Mellons (+1) and OHares (+5). Although the experimental period was shorter
than in the study by OHare, it is suggested that the delayed post-test in Combss
study indicates a more reliable measure of sustained syntactic progress. With both
syntactic maturity scores and overall quality of writing improved, the author
concludes that sentence-combining practice seemed to affect more than syntactic
gains, indeed, gains that were incorporated in what teacher-raters consider improved
quality of writing (p. 321). The correlation between syntactic maturity gains and
overall writing quality is not clearly described, however. Caution is required in
interpreting the results of these papers by Combs, as the trial sample was not
randomized.
Hunt and ODonnells (1970) study, which used a sample of 335 students, was
again a clustered trial without randomization. Its aim was to examine the impact of
sentence-combining on the writing of fourth grade students, specifically with 194
black and 141 white students. Again, the measures used to gauge syntactic maturity
were words per T-unit, clauses per T-unit and words per clause. As in Combss
studies, gains were two grade levels for the experimental groups, with particular
gains in syntactic maturity for black students; but there was no delayed post-test, so
gains might have been short term. In general, this is a study with high validity and
reliability, with a relatively large sample, but constrained by the fact that the pre-test
did not include a writing sample and the fact that there was no delayed post-test.
Finally, the clustered trial (i.e. in existing groups, like classes) nature of the study,
without randomization, means that we cannot be sure that other factors not
mentioned in the study did not bear some influence on the results.
If we look at the effect sizes of the four studies mentioned so farthe four that
provide the best evidence in answering our research questionwe can see that with
regard to the outcome measure of words per T-unit (which is regarded by these
authors as the best measure of syntactic maturity) OHare (1973) finds a very large
positive effect for the intervention of sentence-combining on writing accuracy and
quality (effect size52.4, CI 1.81 to 2.94). In studies by Combs (1976, 1977) this
effect (post-test effect 1.09, CI 0.66 to 1.50) is confirmed, but found to lessen
somewhat as measured by delayed post-test (effect 0.68, CI 0.27 to 1.07). All three
of these results are statistically significant.

Effect of grammar teaching on writing development 51

An overall synthesis of the results from the 18 studies examined in the in-depth
review comes to a clear conclusion: that sentence-combining is an effective means of
improving the syntactic maturity of students in English between the ages of 5 and 16.
All but two of the studies specify the age group they worked with: predominantly,
this group ranged from fourth grade (910-year-olds) to tenth grade (1516year-olds), with the majority clustering in the upper years of primary/elementary
schooling and the lower years of secondary schooling. The differences between the
studies largely inhere in the degree of advance that students learning sentencecombining enjoy in terms of their syntactic maturity. In the most reliable studies,
immediate post-test effects are seen to be positive, with some tempering of the effect
in delayed post-tests. In other words, as might be expected, gains made by being
taught sentence-combining in terms of written composition are greatest immediately
after the intervention, and tail off somewhat thereafter. Significantly, in the one
study that undertakes a delayed post-test, syntactic maturity gains are maintained,
albeit less dramatically than immediately after the event.
Conclusion
The present systematic review of the effect of grammar teaching in English on 516year-olds accuracy and quality in written composition has been the most extensive
undertaken to date, even though we have excluded unpublished Ph.D. theses and
focused only on those studies deemed to be of high to medium quality in our indepth reviews. The quality of seven out of ten of the primary studies included in the
in-depth review on the teaching of syntax is a limitation on the review, as is the lack
of recent research. The 17 studies in the sentence combining review that show a
positive effect were all conducted in the USA; the one that showed no positive effect
was conducted in Canada. There is a question as to how generalizable the results will
be in relation to countries outside the USA. The majority (14) of the studies in the
in-depth review were published in the 1970s and 1980s; two were published in the
1960s, and one each in the 1990s and 2000s.
On the basis of the results of the two in-depth reviews, we can say, first, that the
teaching of syntax (as part of a traditional or transformational/generative approach to
teaching grammar) appears to have no influence on either the accuracy or quality of
written language development for 516-year-olds. This does not mean to say that
there could be no such influence. It simply means that there have been no significant
studies to date that have proved such an effect. This first key point must be qualified
with a caveat. There was considerable difficulty in synthesizing studies on the
teaching of syntax because of their heterogeneity: they used different intervention
materials; different analytical frameworks; there was some methodological invalidity
or unreliability; and the relationship between transformative and generative
approaches, though clear in the linguistics literature, was not clearly articulated
pedagogically.
Second, the teaching of sentence-combining appears to have a more positive effect
on writing quality and accuracy. The studies are relatively more homogeneous,

52 R. Andrews et al.
though they do reveal a difference of degree in the extent to which sentencecombining techniques help to improve writing development.
We do not pretend that our in-depth studies of the teaching of syntax and
sentence-combining cover all aspects of the teaching of grammar. However, we can
say that there appears to be a distinction between the two approaches we have
reviewed. For instance, the teaching of syntax appears to put emphasis on
knowledge about the construction of sentences. Sentence-combining suggests a
pedagogy of applied knowledgeat its best, applied in situations of contextualized
learning; at its worst, drilling. The implications for policy and practice are that:

N
N

we should continue to ask whether the teaching of formal grammar is helpful in

improving young peoples writing (and not be sidetracked by titles like Not whether
but how), taking into account the fact that there has been no clear evidence in the
last hundred years or more that such interventions are helpful;
we should look more closely at techniques of sentence-combining and other
practical approaches, which appearon the evidence to dateto suggest more
positive effects on writing development;
if we think that language awareness and other approaches to knowledge about
language (e.g. rhetorical awareness, a focus on genre) are a useful part of the
primary or secondary curriculum, proponents should specify more clearly what
benefits might be had from such attention, and what this attention might mean in
terms of both pedagogy and measures of writing quality;
a review of the National Literacy Strategy in England, and more broadly, of the
National Curriculum for England, should take place to identify which of the
methods for improving the quality and accuracy of young peoples writing are
most effective. At present, we have an eclectic approach, patchily implemented. It
is unlikely that real advances in written literacy will take place, particularly at the
bottom end of the range in this particular capability, until such research is
undertaken and its results transformed into effective and exciting teaching.

In terms of future research, we hope we have contributed, at the very least, to

clearing the ground. Future researchers may wish to return to questions of the
effectiveness of formal grammar approaches to improving writing, but will do so with
the lessons of previous studies to build on. Future studies will have to be more
rigorous than previous ones if they are to add to knowledge in the field, both in terms
of their research design and implementation and also in terms of their critical
reflection. If they are interested in effectiveness, they should design rigorous
randomized controlled trials to determine such outcomes. We feel there is not only
the chance for future researchers to compare different teaching approaches in welldesigned studies, but also to explore more extensively the possibilities for sentencecombining and its associated techniques.
However, we are also aware that our own studies are limited in their range and do
not take into account all the contextual factors that might bear upon young peoples
success in writing. Accordingly, the Centre for Language Learning Research at York,
in collaboration with Dominic Wyse at the Faculty of Education, Cambridge, is

Effect of grammar teaching on writing development 53

undertaking a series of new primary studies, starting with a national survey of the
teaching of grammar in all primary and secondary schools in England; and then
following it with a subsample of interviews with respondents from a range of socioeconomic contexts. From that point, we will design new approaches to the teaching
of grammar and trial these to find out which are the most effective.
Finally, we do not think that questions of effectiveness are the only research
questions that are of interest in terms of the relationship between grammar and
writing development. We believe that questions pertaining to this relationship need
to be re-theorized in terms of what is meant by knowledge about language and how
grammar (or grammars) might be subsumed under this. Given the changing face of
literate practice, we also need to re-theorize knowledge about language in the light of
the increasing digitization of texts and text-based practice. Such imperatives suggest
the need for a range of research approaches. It terms of the focus of the present
study, further desirable areas for future research would include the examination of
the nature of young peoples writing as it develops through the school years. There is
at least a generation of very good research into emergent grammars that describes
and explains writing development, going back to the work of Britton, Barnes, Rosen,
Wilkinson and others from the 1960s to the 1980s; our hope is that such future and
other related research might be informed by the results we present here.
Acknowledgements
The authors would like to thank Kelly Dickson, Diana Elbourne, Jo Garcia, David
Gough and Katy Sutcliffe at the Evidence for Policy and Practice Information and
Coordinating Centre (EPPI-Centre); the steering committee for the English Review
Group: James Durran, Polly Griffith, Nick McGuinn, Gloria Reid, Peter Taylor and
Ian Watt; and the anonymous peer reviewers of the protocols for each in-depth
review, the final drafts of the reports and of the first draft of the present article.
Note
The full technical reports on both in-depth studies can be found at: http://eppi.ioe.ac.uk/reel

References
Andrews, R. (2005) The place of systematic reviews in educational research, British Journal of
Educational Studies, 53(4), 399416.
Asker, W. (1923) Does knowledge of formal grammar function?, School and Society (27 January),
109111.
Bateman, D. R. & Zidonis, F. J. (1966) The effect of a study of transformational grammar on the
writing of ninth and tenth graders (Urbana, IL, National Council of Teachers of English).
Benfer, M. (1935) Sentence sense in relation to subject and predicate. Unpublished Masters thesis,
University of Iowa.
Boraas, J. (1917) Formal grammar and the practical mastery of English. Unpublished doctoral thesis,
University of Minnesota.
Britton, J. (1983) Shaping at the point of utterance, in: A. Freedman & I. Pringle (Eds) Learning to
write: first language second language (Harlow, Langman), 1319.

54 R. Andrews et al.
Catherwood, C. (1932) A study of the relationship between a knowledge of rules and ability to correct
grammatical errors and between identification of sentences and knowledge of subject and predicate.
Masters thesis, University of Minnesota.
Combs, W. E. (1976) Further effects of sentence-combining practice on writing ability, Research in
the Teaching of English, 10, 137149.
Combs, W. E. (1977) Sentence-combining practice: do gains in judgments of writing quality
persist?, Journal of Educational Research, 70(6), 318321.
Cope, B. & Kalantzis, M. (1993) Introduction: how a genre approach to literacy can transform the
way writing is taught, in: B. Cope & M. Kalantzis (Eds) The powers of literacy: a genre
approach to teaching writing (Pittsburgh, PA, University of Pittsburgh Press), 121.
Department for Education and Employment (2000) Grammar for writing (London, Department
for Education and Employment [book and video]).
Department of Education and Science (1975) A language for life (The Bullock Report) (London,
Her Majestys Stationery Office).
Department of Education and Science (1988) Report of the Committee of Inquiry into the
Teaching of the English Language (The Kingman Report) (London, Her Majestys Stationery
Office).
Elley, W. B., Barham, I. H., Lamb, H. & Wyllie, M. (1975) The role of grammar in a secondary
school curriculum, New Zealand Council for Educational Studies, 10, 2641.
Elley, W. B., Barham, I. H., Lam, H. & Wyllie, M. (1979) The role of grammar in the secondary
school curriculum (Wellington, New Zealand Council for Educational Research).
EPPI-Centre (2002a) Core keywording strategy: data collection for a register of educational research.
Version 0.9.7 (London, EPPI-Centre, Social Science Research Unit).
EPPI-Centre (2002b) EPPI-Reviewer. Version 2.5.2 (London, EPPI-Centre, Social Science
Research Unit).
EPPI-Centre (2002c) Guidelines for extracting data and quality assessing primary studies in educational
research. Version 0.9.7 (London, EPPI-Centre, Social Science Research Unit).
Fogel, H. & Ehri, L. C. (2000) Teaching elementary students who speak black English vernacular
to write in standard English: effects of dialect transformation practice, Contemporary
Educational Psychology, 25(2), 212235.
Hilfman, T. (1970) Can second grade children write more complex sentences?, Elementary English,
47, 209214.
Hudson, R. (1992) Teaching grammar: a guide for the National Curriculum (Oxford, Basil
Blackwell).
Hunt, Kellogg, W. (1965) Grammatical structures written at three grade levels. NCTE Research
Report No. 3 (Champaign, IL, National Council of Teachers of English).
Hunt, K. W. & ODonnell, R. (1970) An elementary school curriculum to develop better writing skills
(Washington, DC, Office of Education, Bureau of Research).
Kitzhaber, A. R. (Ed.) (1968) The Oregon Curriculum: a sequential program in English (New York,
Holt, Rinehart & Winston).
Macaulay, W. J. (1947) The difficulty of grammar, British Journal of Educational Psychology, 17,
153162.
McNeill, J. H. (1994) Instruction for deaf students in syntactic cohesion, Acehi Journal/Revue
Aceda, 20(3), 8895.
Mellon, J. (1969) Transformational sentence combining. Research Report No. 10 (Urbana, IL,
National Council of Teachers of English).
OHare, F. (1973) Sentence combining: improving student writing without formal grammar instruction.
Research Report No 15 (Urbana, IL, National Council of Teachers of English).
Perera, K. (1984) Childrens writing and reading: analysing classroom language (Oxford, Basil
Blackwell).
Qualifications and Curriculum Authority (1998) The grammar papers (London, Qualifications and
Curriculum Authority).

Effect of grammar teaching on writing development 55

Qualifications and Curriculum Authority (1999) Not whether but how (London, Qualifications and
Curriculum Authority).
Rice, J. M. (1903) Educational research: the results of a test in language and English, Forum,
XXXV, 209293, 440457.
Roberts, C. M. & Boggase, B. A. (1992) Non-intrusive grammar in writing, paper presented to the
Annual Conference on Computers and Writing, Indianapolis, 13 May.
Robinson, N. (1960) The relationship between knowledge of English grammar and ability in
English composition, British Journal of Educational Psychology, 30, 184186.
Rousseau, M. K. & Poulson, C. L. (1985) Using Sentence-combining to teach the use of
adjectives in writing to severely behaviorally disordered students. Unpublished research
report, New York, City University of New York (ERIC document number ED342153).
Saddler, B. & Graham, S. (2005) The effects of peer-assisted sentence combining instruction on
the writing performance of more and less skilled young writers, Journal of Educational
Psychology, 97(1), 4354.
Satterfield, J. & Powers, A. (1996) Write on! Journals open to success, Perspectives in Education and
Deafness, 15(2), 25.
Segal, D. & Barr, N. R. (1926) Relation and achievement in formal grammar to achievement in
applied grammar, Educational Research, 14, 401402.
Stock, R. (1980) The effect of teaching sentence patterns on the written sentence structures of
grade two children. Unpublished research report, Manitoba, Canada (ERIC document
number ED208414).
Stone, A. K. & Serwatka, T. S. (1982) Reducing syntactic errors in written responses of a retarded
adolescent through oral patterning, Education and Training in Mental Retardation and
Developmental Disabilities, 17(1), 7174.
Symonds, P. M. (1931) Practice vs. grammar in the learning of correct usage, Journal of
Educational Psychology, 22, 8196.
Thomas, G. & Pring, R. (Eds) (2004) Evidence-based practice in education (Maidenhead, Open
University Press).
Thompson, C. L. & Middleton, M. (1973) Transformational grammar and inductive teaching as
determinants of structurally complex writing, California Journal of Educational Research, 24,
2841.
Wilkinson, A. (1971) The foundations of language (London, Oxford University Press).
Wyse, D. (2001) Grammar for writing? A critical review of empirical evidence, British Journal of
Educational Studies, 49(4), 411427.