Where did you learn this? Determining learner attributions for the sources of learning in a clerkship Divy Ravindranath MD MS and Tamara Gay MD Department of Psychiatry, University of Michigan School of Medicine BACKGROUND RESULTS Third year medical students at the University of Michigan are required to complete a 6 week clerkship in psychiatry. During this clerkship, students learn in a variety of settings. They see psychiatric patients, receive supervision from psychiatry residents and attending physicians, participate in weekly didactic sessions with each didactic session covering one topic area in psychiatry, and study on their own. As is the case at many other medical schools,1 students also complete a USMLE “Shelf” Examination at the end of the clerkship. 68% of questions were answered by knowledge gained from self-study. 18% were answered by knowledge gained from conversations with residents/faculty. 12% were answered by knowledge gained from the didactic series. 6% were answered by knowledge gained from patient encounters. These data were statistically different from random responses (chi-squared value = 1115.554, p < 0.0001; Figure 1). Combining data for self-study and didactic series and comparing against pooled data for the other responses continued to show statistical significance (chi-squared value = 364, p < 0.0001). Moreover, attributions for correct responses did not vary from attributions for incorrect responses (chi-squared value = 3.67, p = 0.299; Figure 2). Figure 1: Learner Attributions for Sources of Learning (Percentages) 1) Conversations with residents/faculty (18%) 2) Didactic Series (12%) 3) Experience with a patient (6%) 4) Self-study (64%) HYPOTHESIS We hypothesized that students see learning for the shelf examination as an independent process from learning during delivery of clinical care during the clerkship. If so, then students would report that they learned material to answer shelf examination style questions from self-study and the didactic series, rather than from other activities in the clerkship. METHODS To test this hypothesis, we designed a survey that paired nine shelf examination-style questions with a question asking students where they acquired the knowledge to answer the shelf-style question. The shelf-style questions were drawn from the nine topic areas covered in the didactic sessions. The attribution question was phrased as a forced choice between one of the four setting for learning identified above. The survey was administered immediately before the examination for each clerkship block in the 2007-2008 academic year. Chi-squared value = 1115.554, d.f. = 3, p < 0.0001 Figure 2: Learner Attributions Separated by Correct vs. Incorrect Answer (Percentages) 70.00% SAMPLE SURVEY ITEM: 1. Which of the following is a defining characteristic of delirium? a. Concomitant infection b. Gradual onset c. Psychotic symptoms d. Waxing/waning attentiveness 60.00% 50.00% Correctly Answered Incorrectly Answered 40.00% 30.00% Which of the following were most useful in helping you answer the above question: 1) Conversations with residents/faculty 2) Didactic Series 3) Experience with a patient 4) Self-study 20.00% 10.00% DISCUSSION These data support the hypothesis that students see learning for answering shelf-style questions as independent from delivery of clinical care. When asked to identify the mechanism of knowledge acquisition most useful in answering a Shelf examination-style multiple choice question, students overwhelmingly selected self study. Using a different survey methodology, Briscoe et al found that students primarily used books to prepare for their end of clerkship examinations.2 Given that the medical school admissions process selects for students who have developed techniques for success on multiple choice tests, it is not surprising that students see test preparation as independent from the “work” of the clerkship. Briscoe et al also found that students asked for “more time to read/study” in order to get ready for the clerkship-end examination. However, more time to study may not result in better test scores.3 Given that time spent reading may trade off with clinical experiences during a clerkship and that these clerkship experiences form the basis for development of professional identity, one could question the value of heavily weighting the Shelf exam in a student’s final clerkship grade. Indeed, schools that use the Shelf examination score assign a median fractional weight of 25% to the examination score.1 There are many limitations to generalization of the present study. The most significant limitation is that the survey was administered immediately before the clerkship end examination, which is a time when students are theoretically steeped in preparation for the test. As such, there may have been a recency effect modulating the students’ survey answers. Moreover, the survey has not been formally validated and the data only represent one class of students at one medical school. 0.00% 1) Conversati ons wi th resi dents/facul ty 2) Di dacti c Seri es 3) Experi ence wi th a pati ent 4) Sel f-study Chi-squared value = 3.67, d.f. = 3, p = 0.299 References 1. Levine RE, Carlson DL, Rosenthal RH, et al. Academic Psychiatry. Spring 2005; 29:1, 52-57. 2. Briscoe GW, Fore-Arcand L, Levine RE, et al. Academic Psychiatry. March-April 2009; 33:2, 120-124. 3. Gerhardt JD, Filipi CJ, Watson P, et al. American Journal of Surgery. February 1999; 177, 132-135. Results were analyzed using chi-squared tests of independence. Acknowledgments: Thanks to our medical students for their participation, University of Michigan Residency This project was determined to be exempt by the University of Michigan IRB. Research/Clinical Scholars Track colleagues for input into this project, and Beth Weihe for administrative support.