Medical Students' Learning Strategies:

Evaluation of First Year Changes

John R. Barker and Joanne P. Olson

University of Mississippi Medical Center, Department of Institutional Research, Jackson, MS 39216-4505, and University of Mississippi Medical Center, Office of Grants and Contracts, Jackson, MS 39216-4505

Research on self-regulated student learning strategies has shown that the strategies students use impact academic outcomes, whether in primary school or undergraduate settings. Little work has been done, however, on the use of learning strategies by professional/graduate students. This study evaluated changes in the use of learning strategies during the first year of medical education at a public university. Students in the first year who had become more intrinsically motivated had higher grades than other students. Scores on the scale for self-efficacy, the belief that one will do well, were significantly, positively related to first-year medical school grades. The implications of these findings for medical education are discussed.

Despite rigorous undergraduate training, many students have difficulty mastering the increased intellectual demands of the first-year curriculum of medical education. Anecdotal evidence from the University of Mississippi Medical Center's Learning Assistance Project, an academic skills counseling program for students, suggests that students who encounter difficulties in medical school often use ineffective, study strategies. This study is one of several projects initiated at our campus to investigate the motivation, skills, and resources that influence the learner's participation in the teaching/learning process.

Findings on the role of the individual learner in the process of learning have been generated by decades of research in cognitive psychology. The work of Albert Bandura (1986) crystallized a theoretical foundation for the study of learners in what he termed a social-cognitive theory. Rather than presume individuals as passive actors shaped simply by a stimulus-response orientation (Weinstein and Mayer, 1986), Bandura indicated that human functioning is more accurately described by the reciprocal interactions of environmental factors, individual cognitions, and behavior. The educational application of Bandura's social-cognitive theory manifested itself in a concept known as self-regulated learning. Self-regulated learning has been defined as the "internalization of learning and task management strategies coupled with the ability to mobilize and maintain them when situations demand" (Corno, 1989, p. 111). Research by Pintrich and his colleagues at Michigan (McKeachie et al., 1985) was among the first of several studies to show that college-level instruction, including elaboration skills and active reading, led to increased scores on tests of students' learning strategies and course grades. Subsequently, Pintrich and his colleagues developed a new test, the Motivated Strategies for Learning Questionnaire (MSLQ) (Pintrich et al., 1991), that indicated other elements of self-regulated learning including student motivation, metacognitive self-regulation, effort regulation, and self-efficacy also correlate with undergraduate course grades.

MATERIALS AND METHODS

The MSLQ (Pintrich et al., 1991) was selected as the instrument to measure the components of self-regulated learning explored in this study. Developed by Pintrich, Smith, Garcia, and McKeachie at the National Center for the Improvement of Postsecondary Teaching and Learning (NCRIPTAL) at the University of Michigan, the MSLQ includes fifteen scales. The fifteen scales are intrinsic motivation, extrinsic motivation, task evaluation, control of learning beliefs, self-efficacy, test anxiety, rehearsal strategies, elaboration strategies, organization strategies, critical thinking, metacognitive self-regulation, time and study environment, effort regulation, peer learning, and help seeking.

Students rate each of 81 items on a Likert scale of one ("not at all like me") to seven ("very much like me"). Scores are constructed for each of the fifteen scales. Factor analysis of the fifteen scales previously established the survey's psychometric integrity (Pintrich et al., 1991). Validation studies have found several of the scales to be significantly correlated with high achievement in undergraduate course work (e.g., Pintrich et al., 1991). No equivalent scale for graduate or professional courses was found in the literature. Permission was obtained from NCRIPTAL to adapt the form for processing by a marked-scan optical reader for use in this study.

All 1994-95 first-year medical students at the University of Mississippi School of Medicine who attended the first day of orientation and who signed a consent letter to participate in the study (n=103) completed the MSLQ. First-year students were given the opportunity to take the MSLQ for a second time late in May 1995 during class time. Sixty-four students agreed to complete the MSLQ for a second time; fifty-nine responses were complete and are reported here. This pretest-posttest design was implemented to identify any changes in learning strategies that occurred during the first year of medical education. Data from the students' permanent records were obtained to identify the students' performance in first-year courses. The scores from individual courses were reduced to a single grade (labeled in this study as M1GPA) by a process of taking a weighted average of the individual course grades. Of the 103 students in the 1994/95 entering class, 15.5% were minority students (fourteen Black students and two Hispanic students) and 28.2% were women. The demographics that constitute the 59 cases for this study were somewhat different from the total cohort: 11.9% were minority students and 33.9% were women.

RESULTS

To evaluate the changes between the first and second administrations of the MSLQ, descriptive statistics and paired t-tests were run on each of the fifteen scales. The general trend for many of the variables was to decrease, possibly indicating artificially inflated scores in the pretest measurement and regression toward the true mean in the posttest. However, the changes were not uniform across each of the variables; differences in pretest and posttest measurements were attributed to student attitudinal and behavioral changes from the beginning of the year until the end.

Table 1 presents Chronbach's alphas, measures of internal consistency, for each of the 15 MSLQ scales from the cohort of medical students in this study, with results from a group of Midwestern undergraduate students documented by Pintrich et al. (1991) included for comparison. The only substantial difference between the two groups shown in Table 1 was the .20 difference on the control of learning beliefs scale. Table 2 shows descriptive statistics and the results of the paired t-test with p-values expressed to the nearest thousandth. Statistically significant differences were found on intrinsic goal orientation, task value, control of learning beliefs, rehearsal learning strategies, metacognitive self-regulation, time and study environment, effort regulation, and peer learning. All of the significant changes were in a negative direction, indicating a decrease in scores at the end of the first year from the students' initial scores. Differences were not statistically significant (p>.05) for the changes in extrinsic goal orientation, self-efficacy, test anxiety, elaboration, organization, critical thinking, or help seeking. The changes in learning strategies indicated in Table 2 suggest that students as a group were less likely to use some of the strategies they had arrived at medical school with than they indicated on the first day of orientation.

For the scales on which no significant difference was found from time 1 to time 2, power analysis indicated that the small sample size was a principal reason statistical significance was not approached by two of the scales: organization learning strategies and help seeking behavior. At a conservative statistical power level of .80 and an alpha of .05, the sample would have had to contain 250 additional subjects to detect statistically significant differences in the two scales. Given the lack of difference on the remaining scales, sample sizes of even 1,000 students would have failed to provide statistical power enough to detect differences at the p<.05 level.

An exploratory evaluation of the impact of the fifteen scales of the MSLQ on the total first-year grades was done using three multiple regression analyses with groups of independent variables (pretest MSLQ scores, posttest MSLQ scores, and change scores calculated by subtracting pretest scores from posttest scores) predicting the first-year grade. Using the R² as the statistic of interest, an important shift in the relationships of the variables with the first-year grade was identified. For the 59 students with pre- and posttest scores, pretest MSLQ scores explained 24% of the variability in the first-year grade. At posttest, the MSLQ scores explained 38% of the first-year grade. But most dramatically, the change scores on the fifteen scales explained 43% of the variability in the first-year grade.

The change in intrinsic goal orientation is one of two variables significantly related to the first-year grade (r=.26, p=.049). The direction of the relationship for this variable indicates that as a whole, intrinsic goal orientation decreased for students in the cohort. For the students whose intrinsic goal orientation increased, however, they also realized a higher first-year grade. Most substantial, the correlation between change in self-efficacy and the first-year grade was statistically significant (r=.48, p=.000), suggesting that students who felt confident that they would do well had higher grades than students who felt less confident in their ability to do well.

Of the second subarea of data from the MSLQ, that of the learning strategies, no statistically significant relationships were detected with the first-year grade. In the third subarea, learning resources management, none of the MSLQ resource variables was significantly related to the first-year grade.

Having uncovered the relationship between first-year grades and changes in intrinsic goal orientation and self-efficacy, one brief analysis was conducted to evaluate possible variables that might be associated with the changes as measured by the MSLQ. The analysis introduced sex and race differences in an effort to explain change scores on the two MSLQ scales significantly related to the first-year grade. Men experienced more change in self-efficacy than women (r=.26, p=.049), but were not different on their change in intrinsic motivation. Furthermore, there was no difference between majority students (e.g., Asian, White) and underrepresented minority students (e.g., Black, Mexican American, Native American, and Mainland Puerto Rican) on the change in self-efficacy. Minority students also experienced no statistically meaningful change in intrinsic goal orientation.

DISCUSSION

Although the initial goal of this study was specifically directed toward learning strategies, the issue of intrinsic motivation and beliefs about one's own abilities to succeed were the significant relationships identified in this study. Decreasing scores in intrinsic motivation may indicate that students may be struggling to find the relevancy of what they are doing in the first year to their plans to be a practicing physician. However, students who do find the relevancy and answer the "why am I doing this?" question with intrinsic interest in learning have higher scores on their first-year medical school grade. Other issues like decreases in scores in rehearsal, metacognitive self-regulation, time and study environment, and effort regulation, might be successfully addressed by resources such as a learning assistance program. Given the results of this study, however, these changes would not appear to be directly significantly related to grades in the first year of medical school at the University of Mississippi Medical Center.

This study indicates that academic administrators and higher education decision makers may want to further look at ways to improve medical students' intrinsic motivation and self-efficacy. Deci and Ryan (1987) suggested that intrinsic motivation is driven by an inherent need to be self-determining rather than externally controlled. On the whole, the class declined significantly in intrinsic motivation, indicating a general student trend away from participating in medical school for the internal goal of learning. Focus groups conducted with first year medical students at the University of Mississippi School of Medicine confirm a feeling of disenchantment that comes during the first year of medical school when difficult lecture-based courses and minimal clinical experience tend to resemble students' undergraduate experience. As such, movement away from an intrinsic goal orientation means that students become more test- and grade-oriented; the competitive nature of medical school and the importance of grades to residency selection may drive this process. However, as this study suggests, students who do find affinity for the medical school experience and enjoy the learning opportunity for its own sake, rather than for some extrinsic reward, actually reap the reward of better grades.

Additional study is necessary to make sense of the impact of self-efficacy on student achievement. Although the importance of self-efficacy is highlighted in the work of Bandura (1986) and others, it is impossible here to disentangle the issue of cause and effect from the relationship of self-efficacy and the first-year grade. Specifically, the difficulty comes from not knowing if the students whose performance had been good in the past were influenced to respond positively to the self-efficacy items or whether positive self-efficacy influenced future performance at the end of the year. Both scenarios are probably true. The implication for the classroom is that some students who feel a sense of failure early in their medical school curriculum and who do not experience some successes may ultimately perform poorly in their courses. What this study suggests is that through data collection activities and use of information from student records, it may be possible to identify students who might be at-risk of negative outcomes in medical school.

Finally, this study reinforces the notion that research on student learning has the potential to provide classroom instructors, curriculum designers, and institutional planners with important information for decision making. Using psychometrically sound instruments, involving students and faculty, and using the results seems an appropriate process for guiding the improvement of learning.

LITERATURE CITED

Bandura, A, 1986. Social foundations of thought and action: A social cognitive theory. Prentice Hall, Englewood Cliffs, NJ.

Corno, L., 1989. Self-regulated learning: A volitional analysis. Pages 111­141 in B.J. Zimmerman and D.H. Schunk, eds. Self-regulated learning and academic achievement. Springer-Verlag, NY.

Deci, E.L., and R.M. Ryan, 1987. The support of autonomy and the control of behavior. Journal of Personality and Social Psychology, 53:1024­1037.

McKeachie, W.J., P.R. Pintrich, and Y-G. Lin, 1985. Teaching learning strategies. Educational Psychology, 20:153­160.

Pintrich, P.R., D.A.R Smith, T. Garcia, and W. McKeachie, 1991. A manual for the use of the motivated strategies for learning questionnaire (MSLQ). University of Michigan, National Center for Research to Improve Postsecondary Teaching and Learning, Ann Arbor, MI.

Weinstein, C.E., and R. Mayer, 1986. The teaching of learning strategies. Pages 315­327 in M.C. Wittrock, ed. Handbook of research on teaching, 3rd ed. McMillan, NY.