I was involved in a wee online discussion as part of the TLAD course today. The topic was student attendence in different teaching formats: whether it goes up or down as extensive lecture notes are handed out, or if lecture recordings made available.
For me personally, I’m not sure of a causal relationship between the attendance at a traditional lecture and exam performance. To my mind both attendance and performance are indicative of poor engagement and so it seems to be a complicated picture: according to Arulampalam et al, there is an impact of non-attendance overall, a but bigger impact of non-attendance for more able students.
The Marburger study , where students were assigned to classes with and without attendence policies, was interesting on several points. He concluded that there is a statistically significant link between absenteeism and exam performance, but that actual difference in performance between classes with and without attendance policies was marginal. He states “..when the absenteeism gap between the (register) policy and no-(attendance) policy classes was greatest, a student in the no-policy class was only 2 percent more likely to respond incorrectly to the average multiple-choice test question than was a student in the policy class.” (parenthesis added). In fact, individual students who missed a class still got the answers right 86-91% of the time. In terms of the impact on attendance numbers with and without an attendance policy, it was found that the absence rate doubled as the semester went on in the class without an attendance policy, but stayed the same in the other class.
A wee look over the literature seems to indicate that the impact of attendance depends on several factors, including format and year of study. For example, in this paper by Foldnes it is suggested that for flipped classrooms “Increasing class attendance by one standard deviation was associated with an increase in mathematics performance of 0.28 standard deviations.”. However, for (what appears to be) a more traditional lecture format with online recording available Nordmann etal found “..no evidence for a negative effect of recording use, or that attendance and recording use were related”, but indicated that the observed positive effects of recording lectures depending on year of study and student ability.
I’ve recently been thinking about student attendance rates in our department because someone suggested to me that a significant proportion of serial non-attenders has a negative impact on students who do attend consistently. I sympathise with that point, but I couldn’t find any literature to back it up so it remains anecdotal.
I an wondering about using a ‘flexi-register’. If we were monitoring a manufacturing process we’d monitor a parameter and only investigate once begins to drift outside acceptable limits. Translating that into an educational environment I’d be thinking that for some small classes I might just ‘headcount’, only circulating a register when it fell below a certain value. But again, there appears to be no evidence to support setting an ‘acceptable level’, or that attendance is actually an important correlating factor in educational outcomes.
Hmmm. None the wiser!
References (where links are not publicly available)
 Daniel R. Marburger (2006) Does Mandatory Attendance Improve Student Performance?, The Journal of Economic Education, 37:2, 148-155,