Medical Education, 1982, 16, 31-38

The ATP 3&a

scale for measuring medical students attitudes to psychiatry

P. B U R R A t , R. K A L I N $ , , P . L E I C H N E R t s , J . J . W A L D R O N t , J . R . H A N D F O R T H ? , F . J . J A R R E T T t A N D I. B. A M A R A t
Department o Psychiatry and f $Department of Psychology, Queens University, Kingston, Ontario, Canada

Summary

Key

The development and validation of a thirty item, Likert-type scale designed to measure medical students attitudes to psychiatry-the ATP-30 (Attitudes Toward Psychiatry-30 items)-are described. We had hoped to demonstrate that attitude to psychiatry was not a unitary matter but an amalgam of attitudes to a number of things to do with psychiatric practice. This hope was not fulfilled, as a unitary dimension was obtained. A positive change in the attitudes of students toward psychiatry was demonstrated in third and fourth medical year students in relation to exposure to psychiatry. Such a change was not demonstrable in two classes of occupational therapy students exposed to a course in psychiatry. The reasons for this difference between medical students and occupational therapy students are discussed-there possibly being important implications here for psychiatric curriculum planning in medical school. Lastly, we have demonstrated that the positive change in attitudes amongst medical students was transient rather than lasting-a matter which most studies of attitude change do not address. In spite of the apparent impermanence of the positive change in attitudes among medical students, there are a number of possible uses to a scale such as the ATP30, and these are discussed.
Correspondence: Dr Prakash Burra, Department of Psychiatry, Queens University, Kingston, Ontario, Canada K7L 3N6. Present address: Department of Psychiatry, The University of Manitoba, Winnipeg, Manitoba, Canada. 0308-01

words: *PSYCHIATRY/edUC; *ATTITUDE OF PERSONNEL; STUDENTS, MEDICAL/*pSyChOl; EDUCATION, MEDICAL, UNDERGRADUATE; OCCUPATlONAL THERAPY/edUC; TEACHlNG/methods; RESIDENCY, MEDICAL; ONTARIO
HEALTH

Introduction

10/82/0100-0031

$02.00 0 1982 Medical Education

We have been impressed, as have most teachers, with the complexity and significance of the relationship between attitude and the matter of learning and the relationship between the attitude to various medical specialties and ultimate career choice. A survey of the literature tended to support our a priori view that attitudes to a number of issues in relation to Psychiatry were often of a particularly strong nature, and, in medical schools, negative in direction (for example, Bruhn & Parsons, 1964). Apart from concern about the possible consequence of such a negative attitude for recruitment into the specialty of Psychiatry, we were concerned about the implications of this state of affairs for future patient management at the level of primary care (Eaton & Goldstein, 1977). A third, and related, concern was about the quality of the undergraduate programme in Psychiatry at Queens and the need in this regard, then, was to find a standardized way of assessing the impact of our courses on the attitudes to psychiatry of our undergraduate medical students. The problem here was not only to measure change (if any)-but also to document the direction of this change and test its stability. This, in line with what Mager (1968) says: Whatever else we do in the way of influencing

32

P. Burra et al.
decision that the construction of a Likert-type scale to measure the attitudes of medical students toward psychiatry was a necessary first s t e p w a s that attempts (as described in the literature) to assess the attitudes of medical students to psychiatry, treated this attitude to psychiatry, by and large, as a global mass, rather than an amalgam of attitudes to a number of facets of psychiatric practice. Through a process of discussion the global mass referred to as attitudes to psychiatry was teased out into eight related attitudinal objects. These were: ( 1) Psychiatric Patients, (2) Psychiatric Illness, (3) Psychiatrists, (4) Psychiatric Knowledge, ( 5 ) Psychiatric Career Choice. (6) Psychiatric Treatment, (7) Psychiatric Institutions, and (8) Psychiatric Teaching. It was our contention that the phrase attitude to psychiatry refers to an attitude to one or more of these attitudinal objects; logically, it is also quite conceivable that any one person might demonstrate attitudes to these objects which could well be mutually contradictory and that this contradictoriness could be of varying degree. Informed by this logic, the construction of the ATP proceeded as follows: Each of four of us (PB, JJW, JRH and FJJ) was allotted two of the eight attitudinal objects listed above and given the task of constructing ten short statements for each of the two objects allotted. Each statement would have to make explicit a certain attitude toward the object. Since these statements could take a positive (for example, Item 5 , in Table I), or negative (for example, Item 8, in Table 1) form, each of the four authors was further instructed that of the ten items needed for each attitudinal object five were to be of the positive kind and five of the negative. The need for this was dictated by a desire to mitigate the effects of a response-set on the part of the respondent. Each of the 80 items so constructed by the four researchers was then discussed by the entire group and the construction modified until there was a clear consensus in the group that the item did indeed refer to the attitudinal object in question and only to that attitudinal object; also, for an item to be acceptable there had to be a clear consensus with regard to its valence (that is, positive or negative). The 80 items so derived were then listed in random fashion by reference to a table of random numbers. Respondents were asked to express their agreement/ disagreement to each item in terms of a five point scale ( I =Agree Strongly, 5 = Disagree Strongly). Positively phrased items (numbers 4, 5 , 9-12, 14, 15,

the student. the least we must strive to achieve is to send him away with favourable rather than unfavourable feelings about the subject or activity we teach. This might well be our minimum, and universal, goal in teaching. A large number of studies in the last 15 years demonstrate a positive change in attitude in medical students in relation to exposure to courses of instruction in psychiatry (e.g. Bakewell et al., 1971; Lau & Offord, 1976). The common thread that runs through all these would seem to be that the student-physician was given a good deal of responsibility and the doctor-patient relationship emphasized. Very little (if anything), however, is said in these studies about the stability of the change achieved-while clearly demonstrating that the change would appear to be in a positive direction and that such a change is associated with early and/or intensive patient contact being provided for the medical student. A notable exception is the report of the evaluation of the Cornell Comprehensive Care and Teaching Program, which study (Kendall & Jones, 1967) demonstrated that an increase in positive attitudes of medical students to matters in the psychosocial sphere was transient and related temporally to teaching in this sphere. This is entirely in line with Rezlers review (1974) of attitude changes during medical school. Running counter to this trend in the literature---of attitudes towards psychosocial matters being generally negative and enduring in medical school-is the finding of enduring attitude change (in a positive direction) toward mental illness in a study by Miller, Lenkoski, and Weinstein (1979). Walton (1967) and Naftulin (1974) provide overviews of the evaluation of medical student attitude change in psychiatric education, chronicle a variety of possible pitfalls, and make it clear that the field is obviously fraught with difficulty. Both papers also review the various instruments that have been used to measure attitudes to psychiatry amongst medical students. On the basis of these reviews and an examination of existing scales such as the Thurstonetype (Anastasi, 1961; Nunally, 1978), the OM1 (Cohen & Struening, 1962) and the Medical Student Attitude Inventory (Rosinsky, 1963), we concluded that no entirely satisfactory scale existed and decided on devising our own.

The ATP 30
One further consideration-which resulted in the

The A TP 30 18, 20, 23, 25, and 27-29) were reversed by subtracting the score from 6. A total score was calculated by adding all item scores. Omissions were prorated. A high score on the scale indicates a positive attitude toward psychiatry. Standardization The 80 item version of the ATP was administered to 189 medical students in their second, third and fourth years of training at Queens University. It was also given to 18 psychiatric residents at Queens and to a Canada-wide sample of 144 psychiatric residents who completed a self-assessment exam in psychiatry in 1978 (Leichner & Kalin, 1980). In addition, 63 students in occupational therapy at Queens completed the ATP 80: these students being chosen because it so happened that they received a series of lectures from members of the Department of Psychiatry; and, they happened to be the only other health professions students receiving systematized contact with psychiatry. In order to assess the stability of the attitude toward psychiatry over time, and to evaluate the effects of training on the attitude, the 30-item version of the ATP (see below) was given to a sample of 52 first year medical students (volunteers from a class of 7 9 , and to the third and fourth year medical students who had taken the ATP 80 during the previous academic year. Final Item Selection Item-total correlations were calculated (with the item removed from the total) in three samples: two consisted of medical students and a third of O.T. students, all from Queens. Items were retained for the final scale if they showed item-total correlations of at least 0.30 in all three samples. Twenty items met this criterion. An additional 10 items were kept which had item-total correlations of at least 0.30 in two of the three samples and 0.15 in the third. A further criterion for item selection was that half of the items had to be positively and the other half negatively phrased. The final Attitudes Toward Psychiatry scale contains 30 items (see Table 1). Additional analyses pertained to the question of whether the items dealing with the eight topics formed meaningful subscales. Positive evidence would consist of correlations between items and subscale totals being higher than between the relevant items and the total from the 80-item scale. This pattern of results did not appear. Correlations be-

33

tween the items and the 80-item total were just as high as those between the items and the subscale totals. This pattern of results suggests that the items on the full version of the questionnaire did not form meaningful subscales according to the eight topics employed in designing the questionnaire. The absence of the eight anticipated subscales does not preclude, of course, the existence of other subscales. In order to examine this possibility, a principal components factor analysis was carried out on the ATP 30 among the medical student sample. The ATP 30 and not the full 80 item version of the questionnaire was selected because we wanted to restrict the analysis to items that consistently measured the attitude (i.e., evaluative reaction) to, and not all sorts of beliefs regarding, psychiatry. This factor analysis yielded a sizeable first factor with a root of 6.49 accounting therefore for 2 1.6% of the total variance (and approximately half the variance due to five factors). The second through fifth factors were considerably smaller, with roots of 2.09, 1.81, 1.49 and 1.30 respectively, suggesting that a one-factor solution is reasonable according to the levelling off criterion (Harman, 1976).

ATP 3 : Basic Results 0

Descriptive statistics regarding the scale are contained in Table 2. As expected, psychiatric residents had the highest (most positive) scores. It should be noted, in fact, that all means were considerably above 90, the logical neutral point of the scale. The mean of psychiatric residents (Queens) is significantly higher than the mean of medical students (t(205)=4.73, P<O.OOI). The means of the two resident samples were virtually identical. Psychiatric residents from Queens are also higher than students in Occupational Therapy (t(79) = 1.176, P<0.05). Students in Occupational Therapy had more favourable attitudes . than medical students (t(250)=5.11, P ~ 0 . 0 0 1 ) As can be seen, the range of scores obtained in all respondent samples, including the sample of psychiatric residents, is quite large. Several measures of the internal consistency and reliability of the final scale appear in Table 3. Itemtotal correlations are in line with those of other attitude scales in the social psychological literature. Correlations between the 30 and 80 item versions of the ATP are very high, suggesting that very little valid information is lost by reducing the scale from 80 to 30 items. The split-half reliability is substantial

34

P. Burra et al.
T A B L E . Attitudes toward psychiatry (APT 30) I

The following questionnaire is designed to ascertain the attitudes ot students toward the field of psychiatry. On the following pages you w i l l lind thirty ( 3 0 ) statements lor which you are asked to indicate your own agreement or &agreement on the IBM answer sheet. There are live possible answer choices for each statement and you are asked to mark only one per statement. The answer categories are a \ follows:

. .

. . . .

A ~ \ t r o n g l )agree

B -agree C -neutral ( n o opinion) D -di\agree E +trongly disagree

Thus. 1 l y 0 u strongly disagree with a question. you will mark in the E space (as shown above). Please use an ordinary lead pencil and if you wish to change an answer. please erase the first answer completely. Try t o ensure that the statement number corresponds with the answer number you are marking Sometimes you may feel a\ though you have had the same item before o n the questionnaire. This will not be the case so please do not try to remember how you checked similar items earlier on. Make each item a separate and independent judgement. Since this 1 5 an attitudinal questionnaire. i t I \ important that your answers be offthe-top-of-the-head rather than deliberately thought out.
P I . E 1 S E A.V.7 WEK -11.1. Q U E S T I O N S I Ps!chiatry is unappealing hecause i t makes s o little UIC of medical training. 2. Psqchiatrist\ talk a lot but do very little. 3. Psychiatric hospitals are little more than prisons. 4. I wjould like to be a psychiatrist. 5 . I t i\ quite easy for me to accept the efficacy of psychotherapy. 6. On the whole. people taking u p psychiatric training are running a w a y from participation in real medicine.

7. Psychiatrists seem to talk about nothing but sex. 8. The practice of psychotherapy basically is fraudulent since there is no strong evidence that it is effective. 9. Psychiatric teaching increases our understanding of medical and surgical patients. 10. T h e majority of students report that their psychiatric undergraduate training has been valuable. 1 I . Psychiatry is a respected branch of medicine. 12. Psychiatric illness deserves at least as much attention as physical illness. 13. Psychiatry has very little scientific information to go on. 14 With the forms of therapy now at hand most psychiatric patients improve. 15. Psychiatrists tend to be at least as stable as the average doctor. 16. Psychiatric treatment causes patients to worry too much about their symptoms. 17. Psychiatrists get less satisfaction from their work than other specialists. 18. It is interesting to try to unravel the cause of a psychiatric illness. I Y . There is very little that psychiatrists can d o for their patients. 20. Psychiatric hospitals have a specific contribution to make to the treatment of the mentally ill. 21. If I were asked what I considered to be the three most exciting medical specialties, psychiatry would be excluded. 22. At times it is hard to think of psychiatrists as equal to other doctors. 23. These days psychiatry is the most important part of the curriculum in medical schools. 24. Psychiatry is so unscientific that even psychiatrists cant agree as to what its basic applied sciences are. 25. I n recent years psychiatric treatment has become quite effective. 26. Most of the so-called facts in psychiatry are really just vague speculations. 27 I i w e listen to them, psychiatric patients are just as human as other people. 28. The practice of psychiatry allows the development of really rewarding relationships with people. 29. Psychiatric patients are often more interesting to work with than other patients. 30. Psychiatry is so amorphous that it cannot really be taught effectively.

T\Ht.t, 2. Descriptive statistics of ATP 30 in four samples Sample 2 4 Year Meds (Queens) Psychiatry residents (Queens) Psychiatry re.;idents (Canada) O.T. Students (Queens)
N

Range

Mean 103.76

SD
10.87 9.60 11.90

189

55 140
102 137

IX

116.33
116.02 I11.67

144 63

58 140

50 133

9.99

and comparable to most attitude scales (Shaw & Wright, 1967). The stability of the measure over time was assessed in several samples. Among first-year medical students. the between-test interval of six weeks consisted of a course in anatomy (control period) without psychiatry related experience. In this sample, the test-

retest correlation attained a very respectable 0.87. In the sample of second, third and fourth-year medical students receiving the ATP 80 in 1977-78 the interval consisted of psychiatry related experience (see below). In this sample the test-retest correlation was lower, but still quite sizeable (r=0.69). The psychiatry related experience affected the attitudes, as shown below, but not always in the same direction. The variable kind of influence therefore lowered the apparent stability of the measure. The stability of the test was assessed over an even longer period of time in two samples of medical students who were tested in successive academic years. In one group, tested during their second and again during their third year, the test-retest correlation was 0.64. In a second group, tested during the

The A TP 30
TABLE Reliability measures of the ATP in medical and O.T. 3. students Reliability Measure Range item-total r Median item-total r r ATP SO-ATP 30 Split-half reliability (Sperman-Brown) Test-retest (control period) Test-retest (treatment period)h Test-retest (2nd-3rd year med.P Test-restest (3rd4th year med.)d
a

35

Med. 0.24-0.57 0.38 0.93 0.90 0.87 0.69 064 0.5 1

O.T. 0.10-0.64 0.38 0.94 0.89 0.67

Test administered over control interval of 6 weeks among 52 first year medical students who received no exposure to Psychiatry in those six weeks. Test administered before and after psychiatry related experience in second, third and fourth year medical students (subsample of N = 156), and O.T. students (subsample N=43). Test prior to training in 2nd year and retest prior to rotation in 3rd year ( N =58). Test prior to rotation in 3rd year and retest prior to clerkship in 4th year ( N =62).

clinical experience with psychiatric patients, (2) general exposure to psychiatry (consisting of a course entitled Psychosocial Aspects of Medicine (P.A.M.) given to second year medical students who have limited direct patient contact consisting primarily of medical, not psychiatric patients), and (3) psychiatryunrelated training (consisting of a course in anatomy for first-year medical students and rotation in paediatrics for a sample of third year medical students). Students with the various training experiences were given the ATP 30 before and after the particular training experience.
Results

third and again during the fourth year, the correlation was 0.5 1.
Training Experiences

One of the major reasons for developing the ATP 30 was to provide an instrument for assessing the influence of psychiatric training on the attitudes to psychiatry held by undergraduate medical students. Since such influence of a course may depend in part on its precise nature, it would be important to know in some detail the psychiatric experience (at Queens) of the various samples of respondents. Space does not permit of this here (but details are available upon request from the senior author). Suffice it to say that: (i) O.T. students in both second and third years received only a series of lectures in psychiatry; (ii) students in first-year medicine (N= 52 of 7 9 , used as a control group, received no psychiatry related training; (iii) exposure to psychiatry in second-year medicine is quite general in the form of a course entitled Psychosocial Aspects of Medicine (P.A.M.)-not a course in psychiatry-where direct patient contact is very limited and is primarily concerned with medical, not psychiatric patients; (iv) medical students in the third and fourth (final) years receive an almost wholly clinical experience (no lectures). We were primarily interested in comparing three types of training: (1) psychiatry rotations (during which third and fourth year medical students receive

The effect of psychiatric training on attitudes toward psychiatry are shown in Table 4. A number of significant changes in attitudes occurred. These changes, however, went in both directions and clearly depended on the type of psychiatric training received. A significant change in a negative direction was observed for first-year medical students (who had no psychiatry related experience between the two test administrations), and for third-year O.T. students (exposed to psychiatry through lectures). No significant change in attitudes occurred among second year medical students (taking a course in psychosocial aspects of medicine), among second-year students in O.T. and in third year medical students in the paediatrics rotation. Very significant improvement in attitudes toward psychiatry took place among third and fourth-year medical students, whose training consisted primarily of clinical aspects of psychiatry with a considerable amount of patient contact.
Discussion

We have demonstrated that the ATP 30 possesses face validity (by virtue of the nature of the items and the manner in which they were constructed). We have also presented some evidence as to the concurrent and construct validity of the test in that: (i) a criterion group of a large number of residents training in psychiatry scored higher on the scale than medical students and occupational therapy students; and, (ii) scores obtained by respondents increased in the direction to be expected following formal exposure to psychiatric training. With regard to this last, it must be noted that the statement holds true only for medical students tested in the third and fourth medical years before and after exposure to rotations in psychiatry. While a similar change was expected in

36

P. Burra et al.
T A H I J Attitudes toward psychlatry among medical students before and after three types of 4. trainine
Time of Measurement Sample and Year of Testing

Before

.After

t-test 5.72s * * 2.29

2.34* 4-2O** * 3 18**

I . Third year Med 77 78 2. Third year Med 78 79 3 . Third year Med 78 79 4. Fourth year Med 77 78 5. Fourth year Med 78 79

Psychiatric Rotations sx 10541 36 104.64

28 10204

47
69

102.70 105.15

I 10 09 I08.5X 106 14 108.06 iny.52

(I.

General Exposure 10 Piychiatr? ( P . A M . ) Sccond )ear Med 77 7X 51 104-38 10492

0.38
- 2 99** -204

Psychiatry-Unrelated Training 7 . F i n t ycar Med ( 4 n a t . ) 78 74 52 101.56 X Third gear Med (Paed.) 78 79 28 105.50

97.98
102.04

* P<0.05. * * P<O.OI * * * P<0.00I. N.B. For the sake of clarity. Occupatlonal Therap? ( 0 T student scores h a \ e heen omltted from this table

occupational therapy students exposed to a series of lectures on psychiatry-no significant increase in scores on the ATP could be demonstrated at the end of the course. This is a somewhat surprising result--which stands out in contrast to the significant increase in scores of medical students. While differences in educational level, and possibly differences in socio-economic status. might contribute to this, we suspect that an important factor in affecting attitudinal change is contact with psychiatric patients and (presumably) with psychiatric expertise in dealing with the clinical problems of these patients. If this suspicion is correct. it would seem that students (including medical students) are more likely to be affected positively when psychiatry can show that it possesses knowledge and skills demonstrably useful in everyday practice-as opposed to theoretical concepts, the practical significance of which is lost on students. If what we surmise is true, there are important implications here for undergraduate psychiatric teaching: the first exposure to psychiatry in medical school-which may well contribute to the basis of a relatively enduring set of attitudes toward this subject-perhaps ought to be one in which the student is exposed to patients manifesting one or more clinical problems and the psychiatrist/teacher demonstrating how these problems are elicited, explored, and treated successfully. This, rather than exposure to theoretical issues.

Mention must be made of an important factor which influences the measurement of attitudes. Walton (1967) has pointed out. in discussing the matter of covert as opposed to overt attitudes, that: a person in confidential conversation may admit views that he would not be prepared to express openly. As a precautionary measure against this, students were, (i) told explicitly that the attitude scale had nothing to do with evaluating them in terms of grades and the like, and, (ii) that they were welcome to use a pseudonym (mothers first and maiden names). It is, of course, impossible to say what effect these measures had in ameliorating anxieties about expressing a negative attitude to psychiatry honestly; or. for that matter. in discouraging a tendency to portray a dishonestly favourable attitude to psychiatry in order to gain favour with the Department. Judging from comments made by medical students in relation to evaluating the courses offered by the Department, we consider it unlikely that either of these two influences exerted a significant effect; it is our experience that the vast majority of students offer, in this context, both positive and negative comments. Secondly. the likelihood that the ATP 30 scores are not affected significantly by the snare pointed to by Walton, is strengthened by the fact of the scores remaining stable over time in the control samples of medical students with no exposure to psychiatry. While we have adduced a certain amount of

The A TP 30

31

evidence for the ATP 30 to be a reliable measure of attitudes both in terms of split-half reliability and in relation to the stability of the measure over time, mention must be made of a small but significant decrease in attitudinal scores in one sample when the scale is filled out by respondents at the beginning and end of a six-week control period during which there is no exposure to psychiatry. This is what obtained in a group of first year medical students. There is a suggestion here that repeat administration of this scale results in a small but definite lowering of the ATP 30 score. Perhaps this is to be understood as a function of irritation on the part of the respondents at being studied. Certainly, some of this sample of students grumbled at having to fill it out again. As for the matter of attitude change we have demonstrated that: (i) Attitude to psychiatry is on the whole quite stable over a period of training unrelated to psychiatry (for example, the paediatrics rotation in the third medical year) or only generally related to psychiatry (for example, Psychosocial Aspects of Medicine in the second medical year). (ii) Attitude to psychiatry is significantly improved as a result of specific patient-oriented training in psychiatry (for example, the third and fourth medical-year rotations in psychiatry). However, and to our disappointment, the more positive attitude to psychiatry ascribed to this period of training would seem to be a transient phenomenon. This, in contrast to the experience of Miller. Lenkoski & Weinstein (1979), may of course only be a reflection of a lack of enduring impact (on attitudes to psychiatry) of the psychiatric training imparted in the third medical-year at Queens. Thus, at the beginning of the fourth medical-year rotation in Psychiatry, the mean score for the groups was found to have returned to the level obtained by the students at the beginning of the third medical-year rotation. We have no reason to believe that the subsequent increase in scores ascribed to the Clinical Clerkship in Psychiatry is again anything but temporary since we have no test data on fourth-year medical students after the lapse of some time following the end of the rotation in Psychiatry. Even so, we would propose that the ATP 30 serves a very useful purpose: the measurement of the attitudinal impact of a course in psychiatry, whether or not such impact is permanent. We have demonstrated apparently differential effects of two sorts of courses-the one almost wholly didactic (for O.T. students) and the other almost

wholly practical (the third and fourth medical-year rotations in psychiatry). It would be of considerable interest to administer the ATP 30 to students at other centres-particularly to medical students exposed to various kinds of courses in psychiatry-and this is in progress (at Queens and Winnipeg, at the moment). It is also quite possible that courses in psychiatry at other centres may well be found to have more permanent positive effect on the students. We would predict that the greater the direct involvement of the students with psychiatric patients, the greater the likelihood of a positive change in attitudes to psychiatry being stable. On the other hand, perhaps significant stability of a more positive attitude (once induced) is not likely to occur without the values related to such a positive attitude being reinforced by non-psychiatric teachers from time to time during the students sojourn through medical school. Finally, it would be of considerable interest to compare two sub-groups of students in medical school-the one scoring high on the ATP 30 and the other low. Given two such smaller groups of students, it should be possible to disentangle some of the factors influencing the development of such attitudes to psychiatry; and perhaps to shed some more light on the matter of factors influencing the choice of psychiatry as a career.

Acknowledgement
This study was conducted with the aid of an Ontario Mental Health Foundation grant to Drs J. J. Waldron, P. Burra, F. J. Jarrett and J. R. Handforth, We are grateful to Mrs Marion Jones for her secretarial assistance.

References
ANASTASI, (1961) Measures of interests and attitudes. In A. Psychological Tesring. 2nd edn. pp. 528-263. McMillan, New York. BAKEWELL. W.E., EWING, J.A., ROWSE.B.A. & COCKRANE, C.M. (197 I ) Experimental clerkship for prefreshman medical students. Journal o Medical Education. 46, 156. f O.A. (1964) Medical student attitudes BRUHN. J.G. & PARSONS, toward four medical specialties. Journal o Medical Education, 39. f 40. COHEN. & STRUENING. (1962) Opinions about mental illness J. E.L. f in the personnel of two large mental hospitals. Journal o Abnormal and Social Psychology, 64. 349. EATON.J.S. JR. & GOLDSTEIN. L.S. (1977) Psychiatry in crisis. American Journal of Psychiorry. 134. 642. HARMAN. H.H. (1976) Modern Facror Analysis. The University of Chicago Press. Chicago. KENDALL. & JONES, P.L. J.A. (1967) General patient care: learning aspects. In Comprehensive Medical Care and Teoching, pp. 77-8.

38

P. Burra et al.
P.yhiatric Education. Edited by H. L. MUSLIN. American Psychiatric Association, Washington. NUNALLY, (1978) Psychometric Theory, 2nd edn, McCraw-Hill, J. Toronto. REZLER,A.G. (1974) Attitude changes during medical school: a review of the literature. Journal o Medical Educafion, 49, 1023. f E.F. ROSINKSY, (1963) Professional, ethical and intellectual attitudes of medical students. Journal of Medical Education, 30, 1016. SHAW, M.E. & WRIGHT, J.M. (1967) Scalesfor the Measurement o f Attitude. McGraw-Hill, New York. WALTON, H.J. (1967) The measurement of medical students attitudes. British Journal of Medical Education, 1, 330.

Edited by < ; r o R G r G. RFADEK M A K Y W. Coss, Cornell & E. University Press. Ithaca, New York. h l l , A.Y H. & O f t O K D , D.R. (1976) A study of student attitudes toward a psychiatric clerkship. Journal of Medical Education. 51, 919. R. LUCHNLK. & KALIN. (1980) Reaulta of the first Canadian P. psychiatric knowledge self-assessment for residents. Canadian Journal oJ Psychiatry, 25, 28 I . MAGFR, R.F. ( 1968) Developing Attitude to Learning. Fearon, Palo Alto. California. MILLER. S.I., LENKOSKI. L.D. & WEINSIEIN. (1979) Enduring D. attitude change in medical students. Journal of Psychiatric Education, 3, I7 I NAPTULIN, D.H. (1974) Difficulties in evaluating medical student attitude change in psychiatric education. I n Evaluative Methods in

Received 25 March 1981; accepted for publication 6 August 198 1