Results table 1
| Study Name | Clarifying the relationship of Medical Education and Moral Development |
| Study Number | 2 17 |
| Author(s) | Donnie J. Self, Margie Olivarez & DeWitt C. Baldwin Jr. |
| Publication date | May 1998 |
| Study Aim | “To assess the development of the moral reasoning skills of medical students”17. |
| Study Design | Longitudinal |
| University | Texas A & M University Health Science Centre College of Medicine |
| Test used | Rest’s Defining Issues Test & A paper and pencil version of Kohlberg’s original Moral Judgement Interview (MJI) |
| Period of time | Over 4 years from the 1991-1994 (inclusive) |
| Intervals | At 3 points during medical school |
| Aspects measured | Moral reasoning |
| Assessment | Quotations from the interviews were used |
| Sample Size | 95 |
| No# of Males | 56 |
| No# of Females | 39 |
| Dropout Rate (%) | Not stated |
| Mean age (years) | Not stated |
| Range of ages (years) | Not stated |
| Ethics teaching | Yes |
Results table 2
| Study Name | Does medical education inhibit the development of moral reasoning in medical students? A cross-sectional study |
| Study Number | 116 |
| Author(s) | Donnie J. Self, Dewitt C. Baldwin, JR. |
| Publication date | October 1998 |
| Study Aim | To investigate whether medical education inhibits the development of moral reasoning in medical students |
| Study Design | Cross-sectional study |
| University | Texas A & M college of Medicine |
| Test used | Defining issues test (DIT) of Rest |
| Period of time | 4 years from 1991 to 1995 |
| Intervals | No intervals were used. Students in years 1-4 were tested at the same point in time. |
| Aspects measured | Moral Reasoning |
| Assessment | Quotations from the interviews were used |
| Sample Size | 851 |
| No# of Males | 511 |
| No# of Females | 340 |
| Dropout Rate (%) | 18.4 |
| Mean age (years) | 24.9 |
| Range of ages (years) | 18-37 |
| Ethics teaching | Yes |
Results table 3
| Study Name | Changes in students’ moral development during medical school: a cohort study |
| Study Number | 318 |
| Author(s) | Johane Patenaude, Theophile Niyonsenga, Diane Fafard |
| Publication date | April 1st 2003 |
| Study Aim | “To investigate the progress in moral reasoning in a cohort of medical students over their first 3 years” 18. |
| Study Design | Longitudinal study |
| University | University of Sherbrooke |
| Test used | Moral Judgement Interview (MJI) |
| Period of time | 3 years |
| Intervals | At the start of their 1st year And the end of their 3rd year |
| Aspects measured | Moral reasoning & Answers to the questionnaires were coded by the stage of moral development |
| Assessment | The French version of Kohlberg’s Moral Judgement Interview (MJI) |
| Sample Size | Quotations from the interviews were used |
| No# of Males | 16 |
| No# of Females | 38 |
| Dropout Rate (%) | Not stated in the study |
| Mean age (years) | Combined=21years & At the end of the 3rd year=21.3 |
| Range of ages (years) | At the end of the 3rd year = 21.3 ± 1.6 Standard Deviations(SD) |
| Ethics teaching | None |
| Study Name | The moral development of medical students: a pilot study of the possible influence of medical education |
| Study Number | 419 |
| Author(s) | D. J .Self, D. E. Schrader, D. C. Baldwin JR & F. D. Wolinsky |
| Publication date | January 1993 |
| Study Aim | “To examine the influence of medical education on the moral reasoning and development of medical students” 19. |
| Study Design | Longitudinal |
| University | Texas A & M University |
| Test used | Kohlberg’s Moral Judgement Interview (MJI) |
| Period of time | Over 4 years |
| Intervals | At the beginning of their 1st year and at the end of their 4th year (upon completion of their medical course) |
| Aspects measured | Moral reasoning |
| Assessment | Quotations from the interviews were used |
| Sample Size | 20 |
| No# of Males | 7 |
| No# of Females | 13 |
| Dropout Rate (%) | 9.1% |
| Mean age (years) | 27.3 , Males average age=27, Females average age=27.5(to one decimal place) |
| Range of ages (years) | 24-35 |
| Ethics teaching | None |
Results table 4
Discussion
All four studies addressed the issue to be studied clearly16-9. The aims of the studies are written in their corresponding table.
The studies were focused with regards to the population being evaluated (medical students), the outcomes (moral reasoning scores) and risk factors (for example, observing doctors practice unethically).
Longitudinal studies are normally used in psychological developmental studies. Observations are taken more than once over a period of time, therefore it is seen as a more reliable measure of change. Longitudinal studies were used in studies two17, three18 and four19 to address their study question. A cross-sectional analysis was used in study one16. Information was collected at specific points in time from different groups: in this case moral reasoning scores of students from years one to four. It was inappropriate to do this, as the aim of the study was to test the changes in the same set of student’s moral reasoning scores during university. Advantages and disadvantages of these types of studies are discussed elsewhere16,19,21-4.
Participants were not selected at random in any of the studies. The cohort was not representative as participants were volunteers, contributing to selection bias and masking the true results.
Reliable and objective tests: the DIT and MJI were used to measure moral reasoning scores; they both reflected what I wanted. It is not possible to determine the exposure bias as we cannot measure the experiences these students have had which help them to mature or develop morally.
Blinding is important as students or interviewers may feel inclined to alter their answers. Single blinding was only mentioned and used in study one16 as participants were not made aware of the purpose of the test during their medical ethics discussions. It is therefore assumed that blinding was not used in the other 3 studies.
Confounding factors such as students being involved receiving ethical education, demographic, cultural and social differences were also identified amongst the studies.
The follow-up used in all the studies was fair and complete enough; the same method was used as in the first and last instance. The follow-up of students should have been at a reasonable amount of time so that any good or bad effects of medical training could have been identified and attributed to a cause.
Drop-out rates are also important to consider as those students lost to follow-up may have scored differently. No reasons were stated in any of the studies for students leaving the study, apart from unreasonable answers being eliminated from the results.
Studies one16 and four19 showed virtually no change in moral reasoning scores across the four years. A difference of less than two points was achieved in study one16. Study four19 also demonstrated no significant changes in moral reasoning.
Weighted average scores used in study one16 narrowed over the four years from 167 (roughly 1.5 developmental stages) to 113(approximately 1 stage).
The mean change found in weighted average scores however, was not found to be significant, although a positive change was shown as p=0.86. This suggests decreased variation among student’s scores; this was matched by the homogeneity of the results in study one16.
The weighted average scores used in study three18 show a decline in moral development at the probability level of p=0.028. The breakdown of results showed that the majority of the sample remained at the same level of moral reasoning (39/54), whilst some regressed (7/54) or progressed (8/54) to a different level. The overall mean change in study three18 wasn’t significant, an increase of 0.02 points from a possible 95 were obtained from first to third year.
According to study three18 a decrease in moral reasoning scores may be due to more frequent use of arguments related self interest or the immediate environment (levels one, two and three of Kohlberg’s theory of moral development). In addition to this, societal perspective was used less (levels four or five of the developmental theory) further contributing to a decrease in moral reasoning scores.
Participants in study one16 and three18 received medical ethics teaching. However, the results of study one16 showed very little difference in their scores over the four years regardless of this teaching. Although, it still may have affected the results. In study two17 scores increased by six points on a scale of 0 to 95 between after a semester of medical ethics teaching. This change was almost twice the amount of the change between the second and last test (2.8 points) although, the length of time between the second and third test was greater. Perhaps the impact of the medical ethics discussions is being shown in these results.
Study one16 was found to be representative in terms of age and gender distribution, as it was comparable to other non-participating medical students. The other studies fail to mention how representative their samples were. Women scored higher than their male counterparts in studies one16, two17 and four19. Study three18 however, showed no relationship between female’s achieving higher scores.
The differences in the rates of change were not found to be significant for those in study four 19and males in study three18. Correlation between age and moral reasoning scores was shown in studies two and four19. No relationship was found in study one16 and the link between gender and scores was difficult to determine in study three18.
P-values were used in the studies two17, three18 and four19 as an indication of whether the studies were accurate. The p-values between periods of testing were measured to see if they where significant.
I don’t believe the full findings of the results because I feel there were too many confounding factors, many of which would be difficult to control over a long period of time.
Three of the studies were conducted at same university and by the same author so perhaps this contributed to bias. The only Bradford-Hill criteria which could be applied to this case would be the strength of association, consistency, specifity and coherence of a study.
The results of study one16 could be applied to the local population as this sample was found are representative. Although, the samples of studies two17, three18 and four19 were not, therefore they could not be fairly applied to their local population.
The combined results suggest that moral development doesn’t progress with medical training. However, this may not be the case as results as extreme results maybe made even when results are combined; therefore this will not reflect any change in the results.
Randomization of participants was not used in any of the four studies16-9. This creates potential selection bias. The people involved in the sample are those who were interested enough to participate. In studies two17 and four19 students participated voluntarily and informed voluntary consent was also given. Studies one16 and four19 however, do not mention whether participation was voluntary or informed. A complementary lunch was provided for those in studies one16 and two17, as an incentive to take part17. The results may have been affected by this.
Approval from an institutional review board was gained in studies one16, two17 and four19.
The DIT was used to assess the moral reasoning of students in studies one16 and two17, whilst the MJI was used in studies three18 and four19. The differences between these two tests are shown in the table below.
As the studies were conducted at different universities which have slightly different course structures. There may be bias in that some universities may equip students to be able to deal with ethical dilemmas more so than others.
In the ethics class in study two17 the subject of the DIT was not discussed. Although, one of the objectives of the programme was to apply the principle of justice when solving moral dilemmas; it directly targets moral reasoning skills, and would help the students in solving the ethical conflicts in the DIT or MJI. The results of studies one16 and two17 have shown how medical ethics discussion can aid medical student’s moral development by increasing moral reasoning scores.
Students may have given more honest answers as they weren’t aware of the tests purpose. It is unsure whether students in study one16 knew about why the tests were being conducted. This makes comparing the changes in moral reasoning scores against those of study three18 and four19 difficult as they didn’t receive any ethical guidance.
The sample size in any study should ideally be large enough so that reliable and representative results are obtained and valid conclusions can be drawn. In study one16 the sample size was quite a large: 598 as data was obtained over four years. In descending order; study three18 was next with 95 participants, and then study two17 with a sample of 54. Study three18 had 95 students and study four19 consisted of 20 subjects. The reason for the small sample size in study four19 is that it is a pilot study; a preliminary study which is conducted on a small scale as a trial run.
In study three18, 54 of 92 students from the year group were used; this seems a reasonable although not the ideal proportion of the year group. Study four19 represented a small proportion of the class: 41.7%, this again maybe due to it being a pilot study. Study one16 used 598 students of 851, this is 70.2%. Study two17 had 95 participants although it doesn’t state how much of the year group or university the sample represents.
Studies two17 and three18 didn’t include any results from the people who left the study. The dropout rate in study four19 was 9.1%, whereas study one16 had a high drop out rate of 18.4%. Drop out rates or exclusion of some data wasn’t mentioned in studies two17 and three18, this may suggest that they don’t want the reliability of the results to be questioned.
Kohlberg suggested his theory was gender biased towards men which gives even more reason for the number of males and females used to be equal. Study three had the largest amount of women (79%) and therefore the lowest amount of men (21%). An over-representation of women was also found in study three18. Study two17 had the smallest percentage of women (41.1%) and therefore the highest percentage of men (58.9%). In study one16 60.9% of the sample was made up of women, and men made up 39.1%. Men made up 35% sample of the study four whilst 65% were women. Results from studies indicate that women are more likely to do better and participate in moral reasoning tests. This creates further selection bias. However, it may be the case there are more female medical students, so including more women in the sample would be more representative of the population of medical students.
In studies one16 and two17 students were tested once more than the students in three18 and four19 were only tested twice. Testing moral development during medical school, two or three times is not enough to reach a reliable conclusion. More frequent and equal measurements of moral reasoning scores should be conducted so that pattern of moral development could have been observed and analysed more closely during medical school.
Study three18 was unsure about the validity of its results as it contradicts the results of the previous two studies the author conducted; students moral reasoning scores tended to show a negative trend. Data was found to be missing randomly for reasons not relevant to the study.
Computerized scoring was used in studies one16, two17 and three18; this eliminated any observer bias. In study four19 however, they state that the method for scoring was seen to be fair throughout the study as the same interviewer was used. He was trained by Kohlberg himself. He had a high reliability of assigning the correct level of the developmental theory of 100% in one whole stage and 75% within a third of a stage was achieved by him19.
According to Kohlberg’s theory as age increases and individuals mature so does their stage of moral development. Study four19 had the highest average age (27.3 years) of participants, followed by study one16 with an average age of 24.9 years. Study three18 came in last with a mean age of around 21 years. The results do show some positive correlation between higher scores and age. The difference between the lowest and highest average age is 6.3 years. This is a big difference for this age group where we would assume the most development occurs, and this affects the results. As participants in Study four19 were older on average we would expect them to be more mature, and have a greater capacity for moral reasoning. The converse may be true for study one16. Significant correlation was found between moral reasoning scores and gender in study one16 and two17.
The range of student’s ages was largest in study 116 (18-37 years). Study four19 then followed (24.9). Study three18 had the least variation in age with most students around 21 years old. Study two17 did not state the average age or its range in students.
The test used in studies one16 and two17 included phrases which were random and had no meaning to exclude results which were dishonest, inconsistent and not taken seriously. Any data which was found to be any of these were excluded from the results, for example, in study one there was a loss of 110 participants due to this to eliminate any biased results.
Conclusion
The results of all four articles16-9 overall show a small increase or levelling effect of moral reasoning scores during medical school.
Having appraised the four articles moral reasoning scores do not decrease as much as I would previously have expected having read articles on this matter. The general consensus is that scores appear to be the same when they are combined. Some individuals may increase or decrease in their level of moral development (where a change is regarded as more than half a stage of the developmental theory), whilst the majority will remain at around the same level.
Feudtner and his colleagues suggested that a “hidden curriculum” exists which inhibits moral development in medical training25. Here students learn and develop their own medical morality from observing doctors, nevertheless, these doctors may make unethical decisions themselves. Alternatively, the results may reflect temporal changes in moral reasoning18. These students will undoubtedly undergo further changes during their careers given that they will be confronted with ethical dilemmas in their working life19. Medical schools have a responsibility to ensure that their students are ethically aware and equipped to deal with such dilemmas.
Ideally moral reasoning scores should increase or at least be maintained at a relatively high level during medical school. I feel that guidance by professionals in this area in addition to the proven small-group discussions could contribute towards this. Engaging students rather than lecturing them on these issues, will also make them responsible for their own ethical development. Perhaps another way to increase moral reasoning scores would be to encourage the use of role models and reflective practice.
I think that medical schools should also aim to instil virtuous qualities in students given that these values are seen to be the essence of making good ethical judgements in medicine27.
References
1. Sensagent Dictionary. http://dictionary.sensagent.com/moral+ development/en-en/ (accessed 5th Dec 2010)
2. Borges NJ, Hartung PJ. Stability of values during medical school. Int J Med Inf 2010; 32:779-781.
3. DJ Self, Schrader DE, Baldwin DC Jr, Wolinsky FD. A pilot study of the relationship of Medical Education and Moral Development. Acad Med October 1991; 66(10):629.
4. Knight JA. Moral growth in medical students. Theor Med Bioeth 1995; 16:265-280. 5. Hafferty FW, Franks R. The hidden curriculum, ethics teaching, and the structure of medical education. Acad Med 1994; 69:861-71:862.
6. Olthius G, Dekkers W. Medical education, palliative care and moral attitude: some objectives and future perspectives. Med Educ 25th February 2003; 37:928-933.
7. Branch WT Jr. Supporting the Moral Development of Medical Students. JGIM July 2000; 15(17): 503-8.
8. Farrow L, Wolf ML. Ideals in Action: The U.S. Schweitzer Fellows Programs. Acad Med June 1998; 73(6):658-661.
9. Feudtner C, Christakis NA. Do clinical clerks suffer ethical erosion? Students’ perceptions of their ethical environment and personal development. Acad Med 1994; 69:670-79.
10. DeWitt DC Jr., Baldwin DC Jr, Daugherty SR, Rowley MD. Unethical and unprofessional conduct observed by residents during their first year of training. Acad Med 1998; 73:1195-200.
11. Self DJ, Baldwin DC Jr, Wolinsky FD. Evaluation of teaching medical ethics by an assessment of moral reasoning. Med Educ 1992:26:178-84.
12. Singer PA. Strengthening the role of ethics in medical education. CMAJ 1st April 2001; 168(7):854-855.
13. Rice ER. Stages of Moral Development. 14th December 2010.
http://www.integratedsociopsychology.net/stages-moral_development.html (accessed 10th December 2010).
14. Patenaude J, Niyosenga T, Fafard D. Changes in the components of moral reasoning during students’ medical education: a pilot study. Med Educ 2003; 37:822-829.
15. Enotes.com Reference. James Rest. http://www.enotes.com/topic/James_Rest (accessed 10th December 2010).
16. Self DJ, Dewitt C, Baldwin Jr. Does medical education inhibit the development of moral reasoning in medical students? A cross-sectional study. Acad Med October 1998; 73(10):S91-3.
17. Self DJ, Olivarex M, Baldwin CD. Clarifying the relationship of Medical Education and Moral Development. Acad Med May 1998; 73(5).
18. Patenaude J, Niyonsenga T, Fafard D. Changes in students’ moral development during medical school: a cohort study. CMAJ 1st April 2003; 168(7):840-4.
19. Self DJ, Schrader DE, Baldwin DC Jr, Wolinsky FD. The moral development of medical students: a pilot study of the possible influence of medical education. Med Educ 1993; 27(3):26-34.
20. Anonymous. Doctors and Patients: flying apart? (Editorial) BMJ 2001; 323:952.
21. Farmer R, Miller D, Lawrenson R. Cohort Studies. In: Farmer R, Miller D, Lawrenson R, eds. Lecture notes on Epidemiology and Public Health Medicine. Blackwell Science 1996: 49-50.
22. Bowers D: Cross-sectional studies. In: Bowers D eds. Medical Statistics from Scratch. Wiley, 2002:66.
23. About.com guide. What is cross-sectional research? http://psychology.about.com/od/cindex/g/cross-sectional.html (accessed 14th December 2010).
24. Evidence-based Dentistry. Study Design III: Cross-sectional studies http://www.nature.com/ebd/journal/v7/n1/full/6400375a.html (accessed 14th December).
25. English V, Romano-Critchley G, Sheather J, Sommerville A. Education and training. In: English V, Romano-Critchley G, Sheather J, Sommerville A eds. Medical Ethics Today. BMJ Publishing Group 1995:665-6.
26. Branch WT Jr. Professional and moral development in medical students: the ethics of caring for patients. American clinical and climatological association 1998; 109:218-30.
27. Campbell A, Gillet G, Jones G. Medical Ethics. In: Campbell A, Gillet G, Jones G eds. Oxford University Press 2001:9.
