Time devoted to anatomical dissection in medical school has been on a steady decline over the last decades,1,2 and in many places active dissection by students has been partly or totally replaced by prosection, the use of cadaveric specimens dissected by others (either staff or other students). This reduction in time and in students' active participation in the dissection process has become a matter of sometimes heated debate.3–5 Significant research into the outcome of cadaver-based versus cadaver-free methods of anatomy teaching is lacking to date. Studies comparing the dissection and the prosection approaches have been largely inconclusive, partly because of methodological issues.6
One problem of such comparisons is that they seem to support a rather simplified notion of what dissection is. In those approaches to the discussion, dissection is often treated as if it were a universal procedure with a uniform learning experience for all participating students. Everybody involved in even the most traditional forms of dissection courses, however, can testify to the fact that the course designs of and educational objectives assigned to active dissection vary substantially among medical schools and even individual teachers. And, even more important, some students do seek more active involvement in the dissection process than others do, and we have little idea of what the motivation for their individual choices might be.
In a first attempt to tackle these issues, we have therefore studied the individual activities of students during “traditional” dissection courses. Our aim was (a) to find out what students actually do during the course and how much time they devote to either active dissection or study of dissected material (prosection), (b) to detect patterns in students' behavior, and (c) to identify possible factors that influence their decision to focus on either active dissection or prosection or, perhaps, to avoid any cadaver contact at all.
We decided to cover the following factors because we hypothesized that they could contribute to a predictive model of individual students' behavior in the dissecting room: motivation, attitude towards dissection as a teaching method, ethical and/or emotional objections towards dissection, course evaluation (student satisfaction), and the student's personality.
Dissection courses for second-semester medical students of the two campuses of Berlin's recently united medical schools were studied during one winter semester (from October 2003 through February 2004) when they were still run using two different course designs (Table 1). Our research had no influence on these course designs. At both campuses, groups of 8 to 10 students were assigned to a table with one human cadaver and were allocated one student assistant who supported supervising faculty (see below). All students were expected to engage in hands-on regional dissection and to study the produced specimens during course time and/or extra study times. Attendance was obligatory. No dissection manual was in general use at either campus, giving the faculty members some freedom for defining an individual dissection schedule and technique.
There were two main differences between campuses. First, at campus A, all body regions were dissected during the 15 weeks of the winter semester that we studied, whereas at campus B the limbs had already been dissected during the preceding semester, and neuroanatomy was deferred to the following semester. All courses were divided into phases according to body regions. The amount of time devoted to any one body region was similar at both campuses.
Second, the number and assignment of faculty members differed between campuses. At campus B, every two tables were supervised by one teacher who aided in the dissection process, explained anatomical details where necessary, and had some freedom in his or her individual course design. At campus A, two independent courses (A1 and A2) were led by one course director each. These were assisted by junior faculty, who were assigned to three to five tables. For organizational reasons and unforeseen staff changes, course A2 had fewer teachers than did course A1 (Table 1).
Dissection of each region was concluded by an oral exam (viva), graded pass/fail, that could be retaken once if necessary. Courses were accompanied by a comparable lecture program of up to eight hours per week at both campuses.
Data acquisition and management
Each course day, short questionnaires were distributed, asking the students to report time (in minutes) spent with given learning activities, and students' motivations that day concerning those activities. In addition, some of the weekly questionnaires were also used to ask all students questions about the student's background, his or her attitudes toward dissection, the student's ethical/emotional concerns regarding dissection, and the student's evaluation of the course. For motivation and other additional questions, students were asked to indicate the extent to which each statement applied to them, using a six-point scale (1 = agree completely, 6 = disagree completely). Ratings on that scale were recoded, with high values standing for high motivation, a positive attitude towards dissection, a strong ethical or emotional objection, and a positive course evaluation, respectively.
We offered the following selection categories, which had been tested in a small pilot study:
- Cadaver related/active dissection: “Carried out the dissection myself.”
- Cadaver related/prosection: “Studied dissected material”; “received explanation of dissected material”; “explained dissected material to others” (Cronbach alpha 0.51).
- Not cadaver related: “Read or otherwise studied”; “studied models”; “listened to talks”; “other” (Cronbach alpha 0.21).
- Breaks or other activities not related to the course.
An additional question asked for time spent in the dissecting room outside course hours. In this way, we were clearly able to identify overestimation in cases where the sum of reported daily activities exceeded the actual course time. In those cases, all activity times of that day were reduced proportionately. In all other cases, no recalculation was done, to avoid excessive “blowing up” of single reported activities (such as dissected arm 45 minutes with no other reported activities for that day). Sums were calculated for each type of activity and for each course day. Because the daily course time was different between the two campuses (270 versus 225 minutes), the time spent in each activity is reported as a percentage of the daily course time minus breaks.
For comparisons between courses and for calculations including data from both campuses, activity values for campus 1 (courses A1 and A2) were computed excluding the phases of limb and brain dissection, because these were not part of course B during the studied semester.
Day-to-day motivation was measured every course day by four to six items. Cronbach alpha was between 0.80 and 0.91 (mean 0.87). Scores were aggregated for each course day and for overall motivation.
The following additional questions were asked for the following parameters:
- Student's personal background: sex, age, and personality traits: 25 items from the Big Five test (Neo-PI-R),7,8 a standard basic inventory that assesses personality traits in five broad dimensions (conscientiousness, agreeableness, neuroticism, openness, and extraversion). Each trait was assessed by five items. After recoding, mean values were aggregated for each trait.
- Attitude towards dissection as a teaching method: Questions about this area were asked both at the start and the end of the course because we supposed that a student's attitude might change with experience. There were six items each; mean value scores were aggregated for both time points (Cronbach alpha 0.85 and 0.87, respectively).
- Ethical and emotional concerns related to dissection: Questions about this area were asked in the middle of the course, as we were more interested in general rather than initial concerns. There were nine items; a mean value was calculated for these nine items with high values meaning high ethical or emotional concerns (Cronbach alpha 0.74).
- Course evaluation: Questions about this area were asked at the end of the course. There were 15 items dealing with content, organization, workload, and atmosphere. A mean value score was calculated (Cronbach alpha 0.85).
Participation in the study was voluntary and had no influence on course grades. Anonymity of acquired data was guaranteed, which precluded attempts at “double checking” self-reported data. Identification of students was by reported matriculation number, which was subsequently removed from the dataset to ensure anonymity. Ethical approval was not required for this study. As a reward and incentive for participation, textbooks sponsored by five medical publishers were allotted among participants.
All data computation and analysis was done using SPSS 13.0 (SPSS, Inc., Chicago, Illinois). Statistical significance of differences was assessed using t tests. Level of significance was set at P = 0.05. To identify aspects that would predict students' participation in active dissection and prosection, two multiple linear regressions were performed. Because Cronbach alpha for all cadaver-unrelated values was 0.21, we excluded this category as an outcome variable from regression analysis for being too heterogeneous. To avoid possible suppression effects between many interdependent variables in multiple regression, we reduced the number of included independent variables by exploratively computing their bivariate correlations with the outcome variables and excluding those that were not significantly correlated.
Response rates and sample characteristics
A total of 4,811 daily questionnaires (out of a potential maximum of 7,020) were returned. We excluded all students who returned fewer than five questionnaires. The remaining analysis is based on 4,696 questionnaires from 309 students, which is 83% of the total student population (371). The median age of these students was 21 years (range 18–41); 64% were women.
The average times devoted to different activities in the anatomy lab (excluding limb and brain dissection; see Methods) are summarized in Table 2 (the percentage for active dissection rises to 38% if limb and brain dissection are included). There were significant differences between the three courses in time devoted to active dissection, prosection, and breaks. Using the t test,
- active dissection: P < .001 for A1 versus A2 and for A2 versus B, P < .03 for A1 versus B;
- prosection: P < .05 for A1 versus A2, P < .001 for A1 versus B, not significant for A2 versus B; and
- breaks: P < .01 for A1 versus A2 and A1 versus B, P < .001 for A2 versus B.
The variability among students in relation to one activity, active dissection, is given as an example in Figure 1. Averaged over the entire course for each student, active dissection ranged from 0% to 82% of daily course time. Other activities showed similar variability: prosection ranged from 0% to 69%, cadaver-unrelated activity from 0% to 78%, and breaks from 0% to 39%.
Average time devoted to activities in the dissecting room outside official course time was 0.9 hours per week (range 0 to 8) for all students (A1: 0.7; A2: 1.0; B: 1.0). It did not correlate with activity levels during course time (r = 0.07 for active dissection, r = 0.12 for prosection). The difference between courses was statistically significant for A1 versus A2 alone (t test, P < .02).
Changes over course time
Average active dissection time declined slowly over the course of the semester in all courses. As an example, data from course A2 are shown in Figure 2. The level of activity depended on the anatomical region to be dissected at each phase of the course. Active involvement was highest for the times of limb dissection (mean 59%, range 7–93, SD 20), followed by the times for dissection of body cavities (thorax and abdomen, mean 33%, range 0–88, SD 20), times for dissection of retroperitoneal space and pelvis (mean 36%, range 0–100, SD 23), and lowest for times for head and neck and/or brain dissection (mean 25%, range 0–89, SD 22). This pattern was similar for all three courses. In complementary fashion, amounts of course time for prosection and cadaver-unrelated activities both slowly increased over the course of the semester. For example, in course A2, prosection took up less than 10% of course time on the first days and around 30% on the final days of the course, with peaks of up to 52% on days preceding exams. Cadaver-unrelated activities rose from around 10% to around 45% of course time in the final week, without obvious peaks.
In addition, time devoted to active dissection was regularly low for days preceding oral exams and for the first day of the abdomen/thorax phase when the body cavities were opened, discussed, and studied in detail. On preexam days, prosection activity peaked, whereas cadaver-unrelated activity and breaks remained largely unchanged.
To assess whether individual student activity preferences were stable over the course of the semester, we analyzed correlation matrices of individual daily values for each activity. Correlations were rarely negative, and the average of correlation coefficients was around r = 0.25 for active dissection, prosection, and cadaver-unrelated activity in all three courses. Correlations for breaks were higher (around r = 0.40).
Because the questionnaires regarding student variables other than motivation were distributed for each question at different time points, the number of respondents answering the different questions varied (Table 3).
On average, students judged themselves as being motivated for each course day (mean 4.43, SD 0.63). Average motivation values showed no statistically significant differences between the three courses. Mean values were fairly stable over the course of the semester. A correlation matrix of individual daily motivation values showed positive correlations for virtually all values, with an average correlation coefficient of r = 0.41 (range −0.03 to 0.71).
Students' attitudes towards dissection as a teaching method were generally positive at the outset of the course (mean 4.83, SD 0.84) and decreased only slightly towards the end (mean 4.71, SD 0.85). Course evaluation was also positive (mean 4.63, SD 0.58).
Ethical and emotional concerns regarding dissection were rated as moderate to low (mean 2.27, SD 0.74). Among all nine items, concerns on religious grounds were lowest (mean 1.47, SD 0.79). The most prevalent reason for concern was uneasy feelings triggered by those parts of the cadaver that are more easily associated with a living person, like the face or hands (mean 3.42, SD 1.55).
Prediction of student activities
To select predictors for multiple linear regression models, bivariate correlations with the two outcome variables were computed. We found several statistically significant, if low, correlations between the surveyed variables and student activities (see Table 3). To exclude chance correlations, we restricted further analysis to statistically significant correlations.
For regression models, both with active dissection and with prosection as outcome variables, motivation, attitude towards dissection (postcourse), course evaluation, ethical/emotional concerns, and conscientiousness were included as independent variables. For active dissection, the model accounted for 9% of the variance. Only one of the variables, attitude towards dissection, predicted significantly for active dissection (standardized beta = 0.26, P < 0.05). For prosection as an outcome variable, the model in this case accounted for 4% of the variance. None of the variables predicted significantly for prosection in this model.
As stated earlier, it was useful to classify cadaver-related activities in two activity preferences that proved largely stable over course time: active hands-on dissection, and prosection, meaning the use of predissected specimens for learning purposes. The latter thus includes all cadaver-related learning activities except active dissection. Cadaver-unrelated activities (e.g., studying models, reading, discussing) proved to be more heterogeneous. On average, students devoted 33% of the course time to active dissection, 27% to prosection, and 31% to cadaver-unrelated activities. We have not found any reports in the literature with which to compare our data.
All three courses were organized into phases devoted to certain body regions. The most stable pattern we found across all three courses was the dependence of the amount of active dissection on these regional phases: dissection of the limbs generated the highest levels of active dissection, dissection of the body cavities generated medium levels, and head, neck, and brain dissection generated the lowest levels. If, as in our case, 8 to 10 students are grouped for the dissection of one cadaver, this pattern can be largely explained by access to body regions: although all 8 to 10 students can keep themselves busy with dissection of the limbs and trunk, only a smaller number of students will be able to dissect the head and neck at the same time. It would be interesting to see whether activity patterns would differ if fewer students were assigned to one cadaver.
Apart from this, we found a high variation of activity levels between the three courses, even if similar phases of the courses were compared. These differences are not easy to account for, because the courses were largely similar in student–cadaver ratios, support by student assistants, additional instruction, etc. Course A2, the course with the lowest levels of active dissection, suffered a shortage of faculty for parts of the course time, which suggests that the level of supervision by faculty and levels of active dissection could be related. Different teachers (course directors as well as junior faculty) will have different expectations and demands, producing different “course atmospheres,” a factor that is, however, difficult to measure. Questions regarding course atmosphere, which were part of our course evaluation, did not prove to be a strong predictor of student activity in our study.
More striking was the surprisingly high variability of specific activities between individual students. Although we found students who reported virtually no active dissection at all, others reported spending more than 80% of their course time with hands-on dissection. Some students obviously managed to spend up to 78% of their course time without cadaver contact.
To better understand what students do during a dissection course and why they obviously seek divergent learning experiences, we had designed our study to look for variables that would influence individual activity patterns. We found several factors that correlate positively with certain activity levels. Most of these factors matched our expectations. Those students with high general motivation, with a positive attitude towards dissection as a teaching method, with higher measures of conscientiousness (as measured by the Big Five test), and who evaluated the course positively, tended to devote more time to active dissection (and less time to either prosection or recreational breaks). Those students with ethical and/or emotional concerns regarding dissection tended to spend less time with active dissection, although not less time with prosection.
Although some of the correlations between personal factors and lab activities were statistically significant, they were not strong, and the predictive model that we generated accounted for only a small part of the variance of students' behavior in the anatomy lab. There may be several reasons for this effect.
First, we are dealing with self-reported data, which causes a considerable measurement error. Although we do not see why students would deliberately misrepresent their activity levels, the subjective estimation of time devoted to a certain activity may, of course, differ from the real time for that activity. In addition, participation rates were high but not complete, because it was difficult to sustain regular student participation over an entire semester. In future studies, some kind of scientific sampling should be considered to improve on this point.
Second, our categories for student activities may not have fully grasped students' individual learning experiences. It may, for example, be possible that one student devotes much time to active dissection because he or she learns a lot this way, whereas another student devotes the same time to active dissection because he or she feels obliged to do so or wants to avoid other activity. As has been discussed in another context,9 it will be important to identify more qualitative facts about students' learning activities to improve our understanding of individual activity choices.
Third, we may have missed some relevant factors. Although we assumed that individual teachers will have an influence on student behavior, we did not ask students to evaluate their individual teachers or to rate the amount of pressure on them (e.g., to participate in the dissection process), simply because we would have lost some teachers' cooperation for our study. The fact that students in the course with the highest devotion to active dissection showed the lowest level of engagement in the dissecting room outside official course hours can be taken as a hint that extrinsic rather than intrinsic motivation played a part in those students' decisions to engage in active dissection. This is, of course, difficult to prove, as is the possible influence of interpersonal dynamics on large-group and small-group levels.
We had also decided at the outset of our study not to include tests of learning style or of visual–spatial ability because we expected that larger test batteries would wear out students' readiness to participate in the study. Moreover, whereas tests for visual–spatial ability are well established,10 the validity of instruments for the study of learning styles is still controversial.11,12
Overall, though, in a given period of an anatomical dissection course, our study suggests that students clearly make very different individual choices and, thus, create their own learning experiences, which are far from homogeneous across the students as a group. Our data suggest that students' preferences for certain activity patterns have at least some stability over the course of the semester. It may, therefore, be attributed in part to students' traits, which are independent of the dissection course itself. This definitely includes personality traits, such as conscientiousness, which we found to have at least a positive correlation with engagement in active dissection, but which may also include learning strategy preferences or visual–spatial ability. Therefore, an attempt to measure the latter in future studies seems warranted.
Two further interpretations of our findings are possible: first, there may be students who avoid active dissection and/or cadaver contact during a dissection course because they belong to a subgroup that does not benefit from these activities at all. If this is true, the size of such a hypothetical group—whether predictors for this lack of benefit can be identified, and whether these students would at all benefit from other ways of learning anatomy—remains to be elucidated by further research. The second possible interpretation, which does not necessarily exclude the first, is that the organizational design of the dissection courses we studied allows for individual strategies in a way that every student can find his or her way to benefit most from the course. If this is true, it would follow that teachers in such courses should not attempt to force a uniform learning experience on their students; instead, they should find ways to accommodate diverse learning preferences.
Our findings confirm that learning anatomy on human cadavers is a complex experience, which is not easily measured. Although our data did not produce a predictive model of student behavior in the anatomy lab, we see our data as a good starting point for future research. One important conclusion from our research is that there is no such thing as the dissection course. Neither courses themselves nor individual learning experiences in such courses are homogeneous. This should always be borne in mind when trying to come to conclusions about dissection as a teaching method.
The authors are particularly grateful to all students who participated in the study. The authors also thank all faculty members for their support of the study, particularly Prof. D. Blottner, Prof. G. Bogusch, and Prof. M. Shakibaei, and the participating student assistants. Finally, the authors are grateful to Dr. M. Knigge and Prof. O. Wilhelm for their statistical support, and to James Ari Liebkowsky for help with the manuscript.
Textbooks and CDs were sponsored by the following German companies: Lehmanns, Börm-Bruckmeier, Schattauer, 3B-Scientific, and Urban&Fischer. These sponsors were not involved in any aspect of the research. Three student assistants were supported by an internal grant from the vice dean of educational affairs of Charité Medical School. The authors did not receive any extra funding.