Medical student research programmes were introduced at Norwegian medical schools in 2002 in order to recruit a greater number of medical students to research. Students in the research programme are intended to conduct independent research under supervision, and to undergo mandatory research training (1). Earlier evaluations have shown that the programme helps to increase the number of medical students who undertake research and to ensure that more students proceed to completing a PhD soon after taking their medical degree (2, 3). It is not known what factors drive students to complete the research programme. As a part of a larger survey among former students on the research programme offered by the University of Bergen, we wanted to investigate what factors had influenced whether students completed the programme and whether they were satisfied with the supervision they had received.
Material and method
The survey was distributed to all medical students who had enrolled in the research programme offered by the University of Bergen since the beginning in 2002 and who had completed their medical degree by June 2017 (n = 149) (2). We established contact with 148 of these people, and 102 (69 %) responded to the survey. Data collection took place in the period 14 November–12 December 2017. The survey included questions about personal background and experiences with the research programme (see Appendix).
A chi-squared test and a Fisher’s exact test were used to compare those who completed the research programme to those who did not, and for factors that were associated with the degree of satisfaction with the supervision received. All data analyses were carried out using STATA version 15.0 (College Station, Texas, USA).
The respondents were informed about the aim of the survey and were told that by submitting the completed questionnaire they were consenting to publishing of the material. The survey was anonymous and the data material was stored on a server with restricted access at the University of Bergen. The study was approved by the Norwegian Centre for Research Data (project reference number 56400).
Results
Out of 102 (69 %) respondents, 88 (86 %) had completed the research programme. The 14 (14 %) respondents who had withdrawn, had attended the programme for a median of 5.5 semesters (interquartile range 5–9). The most frequently cited reason for withdrawing was that the project was not suitable for students on the research programme (n = 8), followed by a problematic relationship with the supervisor or the research group (n = 3). Among those who had completed the programme, 17 (19 %) had considered withdrawing. The most frequently cited reasons were that they had lost interest in the project (n = 8), that the project was not suitable for students in the research programme (n = 6), a problematic relationship with the supervisor or the research group (n = 6), and that the project had a negative impact on their medical studies (n = 6).
Table 1 shows characteristics of the research programme as described by students who completed the programme versus students who withdrew. Students who were not satisfied with their supervision had a lower probability of completing the programme (p = 0.001). Regular supervision (p = 0.088), involvement with other work undertaken by the research group (p = 0.153) and regular meetings (p = 0.267) tended to increase the likelihood of completion. The completion rate did not differ between students who were completely or partly satisfied with the supervision received.
Table 1
Research programme characteristics by student completion or non-completion
|
|
|
Completed the research programme (n (%))
|
|
|
|
Yes (n = 88)
|
No (n = 14)
|
|
Classification of the research programme project
|
|
|
|
|
|
Epidemiological/community medicine
|
|
24 (27)
|
1 (7)
|
|
|
Clinical
|
|
21 (24)
|
4 (29)
|
|
|
Laboratory-based
|
|
41 (47)
|
8 (57)
|
|
|
Other
|
|
2 (2)
|
1 (7)
|
|
Time spent on research, as a weekly average, outside the research programme year
|
|
|
|
|
|
< 5 hours
|
|
21 (24)
|
6 (43)
|
|
|
5–8 hours
|
|
40 (45)
|
6 (43)
|
|
|
> 8 hours
|
|
27 (31)
|
2 (14)
|
|
Research programme workload
|
|
|
|
|
|
Too low
|
|
1 (1)
|
3 (21)
|
|
|
Appropriate
|
|
72 (82)
|
10 (71)
|
|
|
Too high
|
|
15 (17)
|
1 (7)
|
|
Regular supervision
|
|
|
|
|
|
Yes
|
|
48 (55)
|
4 (29)
|
|
|
No
|
|
40 (45)
|
10 (71)
|
|
Supervisor available throughout the study period1
|
|
|
|
|
|
Yes
|
|
81 (93)
|
11 (92)
|
|
|
No
|
|
6 (7)
|
1 (8)
|
|
Supervisor with previous experience of research programme students1
|
|
|
|
|
|
Yes
|
|
37 (43)
|
5 (36)
|
|
|
No
|
|
50 (57)
|
9 (64)
|
|
Involved in other work within the research group
|
|
|
|
|
|
Yes
|
|
51 (58)
|
5 (36)
|
|
|
No
|
|
37 (42)
|
9 (64)
|
|
Regular research group meetings1
|
|
|
|
|
|
Yes
|
|
61 (69)
|
7 (54)
|
|
|
No
|
|
27 (31)
|
6 (46)
|
|
Satisfaction with supervision received
|
|
|
|
|
|
Satisfied
|
|
57 (65)
|
7 (50)
|
|
|
Partly satisfied
|
|
28 (32)
|
2 (14)
|
|
|
Not satisfied
|
|
3 (3)
|
5 (36)
|
A total of 64 (63 %) respondents were satisfied, 30 (29 %) were partly satisfied, and 8 (8 %) were not satisfied with the supervision they received as students on the research programme. Table 2 shows factors relating to the supervision and the research group distributed by the respondents’ satisfaction rate. Students who had received regular supervision and whose supervisor had been available throughout the study period showed the highest rate of satisfaction with the supervision (both p = 0.001).
Table 2
Factors associated with the supervision and the research group by the respondents’ rate of satisfaction with the supervision received on the research programme
|
|
Satisfied
(n (%))
(n = 64)
|
Partly satisfied
(n (%))
(n = 30)
|
Not satisfied
(n (%))
(n = 8)
|
|
Regular supervision
|
|
|
|
|
|
Yes
|
42 (66)
|
10 (33)
|
0 (0)
|
|
|
No
|
22 (34)
|
20 (67)
|
8 (100)
|
|
Supervisor available throughout the study period1
|
|
|
|
|
|
Yes
|
64 (100)
|
23 (79)
|
5 (83)
|
|
|
No
|
0 (0)
|
6 (21)
|
1 (17)
|
|
Supervisor with previous experience of research programme students1
|
|
|
|
|
|
Yes
|
28 (44)
|
11 (38)
|
3 (38)
|
|
|
No
|
36 (56)
|
18 (62)
|
5 (63)
|
|
Involved in other work within the research group
|
|
|
|
|
|
Yes
|
39 (61)
|
15 (50)
|
2 (25)
|
|
|
No
|
25 (39)
|
15 (50)
|
6 (75)
|
|
Regular research group meetings1
|
|
|
|
|
|
Yes
|
44 (70)
|
21 (70)
|
3 (38)
|
|
|
No
|
19 (30)
|
9 (30)
|
5 (63)
|
Discussion
This study investigated factors associated with student completion of the Medical Student Research Programme at the University of Bergen, as well as the students’ degree of satisfaction with the supervision they received. We found a lower probability of completion among students who were dissatisfied with the supervision, compared to those who were partly or completely satisfied. Almost half of those who withdrew from the programme said that part of the reason for doing so was a problematic relationship with their supervisor or research group. The national evaluation of medical student research programmes undertaken in 2006 showed that 75 % of students were satisfied with their supervision while 19 % wanted more supervision (3). We also identified a trend suggesting that regular supervision led to a higher programme completion rate. Many medical faculties have written guidelines that stipulate regular contact between supervisors and students on the research programme (4–6). Furthermore, there was a higher rate of satisfaction among students who had received regular supervision and had access to their supervisor throughout the research period. These findings support the argument that medical students on a research programme need regular follow-up. They also demonstrate the importance of a good relationship between student and supervisor.
The most commonly cited reason for not completing the research programme was that the project was not suitable for a research programme student. This was also one of the most frequently cited reasons for considering whether to withdraw from the programme among students who completed the programme. Properly planned projects that are suitable for medical students with limited knowledge and experience of research are therefore important for keeping students on the programme.
The principal weakness of this study is the reduced statistical robustness caused by the low number of students who withdrew from the programme (2). This may have caused type II errors with true correlations missed. We cannot exclude selection bias despite the high response rate as we had no opportunity to describe the former research programme students who did not respond to the survey.
The assessment of whether projects are suitable for medical students on a research programme, and the students’ relationship with their supervisors, have only been described from the students’ point of view. Nevertheless, the students’ assessment of their time in the research programme is an important perspective which has never previously been systematically investigated and may help to ensure that more students complete the research programme.
The Medical Student Research Programme at the University of Bergen has introduced annual follow-up meetings as an initiative to ensure that students are well supported and that problems are solved as early as possible. Further initiatives to raise the quality of the research programmes offered by Norwegian medical schools may include supervisor training and the introduction of further requirements relating to the supervision and follow-up of students who attend the programmes.
Conclusion
This study shows that a good relationship between student and supervisor, combined with a realistically achievable project, are important factors for ensuring that students complete the research programme offered by medical schools.
