have examined the relationship between student attendance and academic
performance. Durden and Ellis (1995) investigate the link between overall
course grade and self-reported attendance levels in a sample of 346 principles
of economics students over three semesters. Their results, based on OLS
controlling for ability and motivational factors (GPA, college-entrance exam
scores, having had a course in calculus) indicate that attendance matters for
academic performance. In particular, whereas low levels of absenteeism have
little effect on the eventual outcome, excessive absenteeism has a large and
significant effect.
Devadoss and Foltz (1996) examine attendance in a sample of about 400
agricultural economics students at four large U.S. universities. They find
that, even after controlling for both prior grade point average and the degree
of motivation, on average students who attended all classes achieved a full
letter grade higher than students who attended no more than 50 per cent of
the same classes. A positive and significant relationship between attendance
and academic performance is also found by Chan et al. (1997) in a sample
of 71 Principles of Finance students.
More recently, Marburger (2001) investigates the relationship between
absenteeism and exam performance in a sample of 60 students of a principles
of microeconomics course. In this study, information on student attendance
at each class during the semester is matched with records of the class meeting
when the material corresponding to each question was covered. The results
indicate that students who miss class on a given date are significantly more
likely to respond incorrectly to questions relating to material covered that day
than students who were present. Rodgers (2001) finds that attendance has
a small but statistically significant effect on performance in a sample of 167
introductory statistics course. Kirby and McElroy (2003) study the determi-
nants of levels of attendance at lectures and classes and the relationship with
exam performance in a sample of 368 first year economics students, finding
that hours worked and travel time are the main determinants of class atten-
dance, and that the latter, in turn, has a positive and diminishing marginal
effect on grade.
Among studies who reach less robust conclusions about the positive ef-
fect of attendance on performance, Bratti and Staffolani (2002) argue that
estimates of student performance regressions that omit study hours might
be biased, given that hours of study are a significant determinant of lecture
attendance. Using a sample of 371 first-year Economics students they find
that the positive and significant effect of lecture attendance on performance