Critical appraisal

To help you with your critical appraisal skills we are putting "points to note" at the end of all the papers in the studentBMJ. Read and appraise the paper, then check our box to see if you agree with us.

Introduction

About one third of medical students in the United Kingdom add an additional year to the basic five year undergraduate course and intercalate a degree in a medical science (generically a BSc). In some schools intercalated degrees are an integral part of a six year course.

Although many medical teachers believe that intercalated degrees are beneficial, genuine doubt remains about its effect on attitudes and careers. Previous studies have found that medical academics tend to have intercalated a degree,^1,2 and intercalating students may³ or may not⁴ perform better in final examinations. The problem of interpretation was emphasised by the Committee of Vice-Chancellors and Principals' report into clinical academic careers: "the data ...do not (cannot) demonstrate that intercalated degrees cause students to take up academic careers ...it may be that those that are interested in academic research are those that seek to do the intercalated degree."⁵

Although the literature has emphasised a research career as a principal outcome,¹ intercalated degrees may have broader effects. Modern medicine emphasises self directed learning and critical evaluation, skills which may be acquired during an intercalated degree and are relevant to all doctors. This study assesses the effect of an intercalated degree on learning styles and career preferences in a large prospective cohort of UK medical students and examines the effect of the degree in different medical schools.

Participants and methods

In a prospective, longitudinal study of medical student selection and training we studied 6901 applicants in 1990 for admission in 1991 to five UK medical schools.⁶ These applicants represented 71% of all applicants to UK medical schools in that year. We sent questionnaires to applicants with European addresses and received replies from 92% (5361/5845). Applicants were informed that participation was not compulsory and was independent of the selection process.

Of the cohort, 3333 were admitted to any of the UK medical schools to which they had applied (that is, not only the five schools in the selection survey), 2961 in 1991 and 372 in 1992. In 1995 and 1996 all UK medical schools provided information on the progress of the 3333 entrants. A total of 3048 students had entered clinical courses, and 2695 were due to qualify in 1996 or 1997 and formed the subjects for the present study. The students were sent questionnaires about 3 months before final examinations (in 1996 or 1997). Response rate was 56% (1495/2695).

Learning styles (study habits) were assessed by an 18 item version of Biggs's study process questionnaire,^7-10 which has surface, deep, and strategic scales (box).¹¹ Reliability coefficients ( ) were 0.534, 0.721, and 0.637 in applicants and 0.591, 0.734, and 0.727 in final year students.

Career preferences were assessed by a questionnaire rating 27 specialty areas on a five point scale from "definite intention to go into this" through to "definite intention not to go into this," scored 5 to 1.¹² Previous factor analyses suggested seven factors (see table 2), for which mean scores were calculated. The item on medical research was also analysed separately. Academic achievement was coded as average A level grade (A=5; B=4; C=3; D=2; E=1; O/F=0) and number of A levels.

The proportion of students taking intercalated degrees differs between medical schools (a "compositional variable"¹³). We used multilevel modelling to assess the effect of this factor using final year strategic learning score as the response variable, allowing random variation at student and medical school level, and fixed effects of strategic learning at application, A level attainment, the taking of a BSc, the proportion of students at a school taking a BSc, and the interaction of the last two measures.

We used spss for windows version 8.0 for conventional statistical analysis and MLn for multilevel modelling.^13,14 Missing values, which represented about 1% of the questionnaire responses, were replaced by means when appropriate. Denominators are not always equal because of missing values. Significance tests from multiple regression and multilevel modelling are reported as z statistics (estimate/standard error).
 

Results

Intercalated degrees were taken by 904/2695 (33.5%) of the students. Degree classes were known for 795 students: 166 (20.9%) gained a first, 532 (66.9%) a 2.1, 86 (10.8%) a 2.2, and 11 (1.4%) a third, pass degree, or fail.

Learning styles

Students who subsequently took intercalated degrees had significantly lower surface learning scores at application to medical school and significantly higher A level grades and number of A levels than those who did not (table 1). Final year students who had taken intercalated degrees had higher deep and strategic learning scores (table 1). After scores at application were partialled out, final year students who had taken an intercalated degree had higher deep (z=3.73, P=0.001) and strategic (z=4.56; P=0.001) scores (but not lower surface scores (z=0.546, P=0.585)) than those who had not taken an intercalated degree. Significance remained similar after A level results were taken into account.

The average proportion of students taking an intercalated degree at a medical school was 36% (n=28; SD 29%; median 26%; interquartile range 12-51%; range 2-100%). The interaction between the effect of taking a BSc on strategic learning and the proportion of students taking a BSc was significant (z=2.47, P=0.014), the effect of the BSc being greater in schools where fewer students took it. Regression on proportion of students intercalating a degree was significantly negative in those taking a BSc (z=2.57, P=0.010) but not significant in those not taking a BSc (z=0.79, NS).
 

Career preferences

At application to medical school, students who subsequently took an intercalated degree had a higher preference for laboratory medicine and a lower preference for general practice (table 2) and were more interested in medical research (tables 2 and 3). Final year medical students who had taken an intercalated degree had higher preferences for laboratory medicine and medical research and lower preferences for general practice, the effect remaining significant after scores at entry and A levels were partialled out. Students gaining higher degree classes had a greater interest in medical research (z=7.98, P=0.001) and laboratory medicine (z=7.31, P=0.001) and a decreased interest in general practice (z=-3.32, P=0.001) after scores at application were partialled out.
 

Analysis of difference between medical schools was restricted to interest in medical research. Multilevel modelling showed that after research career at application to medical school, mean A level grade, and number of A levels were partialled out there was a significant interaction between the effect of taking an intercalated degree and the proportion of students in each medical school taking an intercalated degree (z=2.60, P=0.010); a significant negative association was seen in those taking an intercalated degree (z=-3.65, P=0.001) and no association in those not taking a degree (z=0.27, NS).
 

Relation between study habits and career preferences

Since taking an intercalated BSc affects both study habits and career preferences, it is important to ask if the effects are independent or mediated.¹⁵ For individual students, regression of strategic learning on taking an intercalated degree was significant after strategic learning at application and interest in medical research at application and final year were partialled out (z=3.95, P=0.001); similarly, an intercalated degree was significantly related to interest in medical research, after interest in medical research at application and strategic learning at application and final year were partialled out (z=7.20, P=0.001). At medical school level, the proportion of students taking a BSc remained significant after covarying the other variable (effect on strategic learning, z=2.12, P=0.034; effect on career in medical research, z=3.34, P=0.0001). Intercalated degrees therefore have independent effects on study habits and career preferences at student and medical school level.

Discussion

This study provides evidence that students taking an intercalated degree are more interested in medical research and also favour deep and strategic learning. The benefits of the intercalated degree were present 3 years after it had been taken and might be expected to last much longer. Because our study is longitudinal, the hypothesis that the effects of an intercalated degree are due to self selection⁵ can largely be discounted. Although detailed results cannot be presented here, there was no evidence that final year respondents were substantively different from non-respondents based on scores at application. This was also found in our previous studies.¹⁶
 

Differences between medical schools

A simple reading of our overall data might suggest that all medical students should take an intercalated degree. Multilevel modelling, however, shows that as more students in a medical school take an intercalated degree the benefit decreases. The mechanism of this effect cannot be elucidated from our data, but a possible explanation is dilution of resources: as proportionately more students take an intercalated degree there are fewer resources for each student, each member of staff supervising more students. If this hypothesis is correct, proper resourcing of intercalated degrees is necessary for them to be effective.

Direction of effects

It might be argued that our study does not show positive effects of the intercalated degree, but rather that the degree mitigates the negative effects of the rest of the course, preventing a fall in deep and strategic learning and decreased interest in research. That is possible. The effects of the other five years of the course are, however, more difficult to study, since all students take all components of it, and the effects are heavily confounded by maturational and age related changes.¹⁷

Most of the students in our study were taking a traditional curriculum, but many medical schools are now introducing curriculums containing problem based learning.¹⁸ These may themselves increase deep and strategic learning.^19-22 Nevertheless, the intercalated degree remains an option in most new curriculums. Our study provides a baseline from which to determine whether new curriculums will increase deep and strategic learning in their own right.

Contributors: The study was designed by ICM and PR. ICM and BCW were responsible for data collection and analyis. ICM wrote the first draft of the manuscript and all authors contributed to revising the manuscript. ICM is the study guarantor.
Competing interests: None declared.

I C McManus, professor of psychology and medical education

B C Winder, research fellow, Hughes Hall, Cambridge CB1 2EW
>

P Richards, president
>

studentBMJ 1999;07:394-436 November ISSN 0966-6494

1. Evered DC, Anderson J, Griggs P, Wakeford R. The correlates of research success. BMJ 1987;295:241-6.
2. Eaton DG, Thong YH. The bachelor of medical science research degree as a start for clinician-scientists. Med Educ 1985;19:445-51.
3. Wyllie AH, Currie AR. The Edinburgh intercalated honours BSc in pathology: evaluation of selection methods, undergraduate performance, and postgraduate career. BMJ 1986;292:1646-8.
4. Tait N, Marshall T. Is an intercalated BSc degree associated with higher marks in examinations during the clinical years? Med Educ 1995;29:216-9.
5. Committee of Vice-Chancellors and Principals. Clinical academic careers: report of an independent task force. London: CVCP, 1997.
6. McManus IC, Richards P, Winder BC, Sproston KA, Styles V. Medical school applicants from ethnic minorities: identifying if and when they are disadvantaged. BMJ 1995;310:496-500.
7. Biggs JB. Student approaches to learning and studying. Melbourne: Australian Council for Educational Research, 1987.
8. Biggs JB. Study process questionnaire: manual. Melbourne: Australian Council for Educational Research, 1987.
9. O'Neil MJ, Child D. Biggs' SPQ: a British study of its internal structure. Br J Educ Psychol 1984;54:228-34.
10. Biggs JB. What do inventories of students\' learning processes really measure? A theoretical review and clarification. Br J Educ Psychol 1993;63:3-19.
11. Newble DI, Entwistle NJ. Learning styles and approaches: implications for medical education. Med Educ 1986;20:162-75.
12. Crimlisk H, McManus IC. The effect of personal illness experience on career preference in medical students. Med Educ 1987;21:464-7.
13. Goldstein H. Multilevel statistical models. London: Arnold, 1995.
14. Woodhouse G. Multilevel modelling applications: a guide for users of MLn. London: Multilevel Models Project, Institute of Education, 1996.
15. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic and statistical considerations. J Personality Soc Psychol 1986;51:1173-82.
16. McManus IC, Richards P, Winder BC, Sproston KA. Clinical experience, performance in final examinations, and learning style in medical students: prospective study. BMJ 1998;316:345-50.
17. McManus IC. Medical students: origins, selection, attitudes and culture [MD thesis]. London: University of London, 1985.
18. General Medical Council. Tomorrow's doctors: recommendations on undergraduate medical education. London: GMC, 1993:1-28.
19. Coles CR. Differences between conventional and problem-based curricula in their students' approaches to studying. Med Educ 1985;19:308-9.
20. Vernon DTA, Blake RL. Does problem-based learning work? A meta-analysis of evaluative research. Acad Med 1993;68:550-63.
21. Albanese MA, Mitchell S. Problem-based learning: a review of literature on its outcomes and implementation issues. Acad Med 1993;68:52-81.
22. Måtenson DF. Students' approaches to studying in four medical schools. Med Educ 1986;20:532-4.