Attendance and Achievement in Medicine: Investigating the Impact of Attendance Policies on Academic Performance of Medical Students

Background: The attendance mandate for the medical course in Melaka Manipal Medical College, Manipal, India was increased from 75% to 90% based on the assumption that the mandatory increase will improve the students' performance. Aims: To find out whether there is any correlation between class attendance and academic performance. Subjects and Methods: This was an institution based retrospective analytical study. Students who have completed Phase I (first two and a half years) of the MBBS course were included in the study. Student marks and attendance, from the database were obtained from three random batches, each, from two clusters A and B respectively. Those who had a mandatory attendance requirement of 75% belonged to A (n = 243), and those who had a mandatory attendance percentage of 90% belonged to B (n = 360). Statistical analyses performed included, Pearson 2 tailed correlation to correlate class attendance with student performance; Cluster analysis to classify group average in a similarity matrix; t-test to determine significance of difference in percentage of students who attained 100% when the college changed mandatory attendance from 75% to 90%; Mann-Whitney test to find out if there was a better performance in university exam when attendance policy changed. Results: There was a significant correlation between attendance and the students who passed in the University exam. The number of students in the pass category was maximum (>90%) compared to students in distinction and failed categories. Percentage of students with 100% attendance rose from 4% (n = 10) to 11% (n = 40) when the mandatory attendance was increased from 75% to 90%. Conclusions: Attendance policy correlated with better academic performance. Reducing absenteeism, probably contributed to the improved academic performance of the students. But the link between attendance and best and worst performances could not be predicted because of small numbers in every batch.