Behavioural Economics, Technology and Mandatory Attendance Policies for College Students

Attendance is voluntary in many college classes, primarily because of the difficulty in taking attendance on a regular basis, but also because of the view that students should have some autonomy in determining the manner in which they engage with academic material. Mandatory attendance policy becomes more of an issue, however, where there is a 'professional' element to a programme. In nursing, for example, there is a high minimum attendance stipulated by the Irish Nursing Board (2005): students must attend 80% of a minimum of 1,533 hours.

There may also be outcomes other than academic achievement which have an important relationship with students’ lecture attendance. A recent report from the institutional research office at UCD has recommended that mandatory attendance policy could be one mechanism to reduce student drop-out (Blaney and Mulkeen, 2009): 'Student Retention in a Modular World - A Study of Retention of UCD Entrants: 1999-2007'. In addition, I recently mentioned on the blog that Irish colleges may soon be offered financial 'incentives' to meet targets in areas such as the retention of students and the rate of course completion. "If they fail to meet these targets, they will face financial penalties." So there are plenty of reasons to care about students' attendance.

There is much debate on what incentives or penalties are appropriate in relation to mandatory attendance policies. This is because penalising students for not showing up can be seen as double jeopardy: they would be punished by lower test scores in addition to a lower attendance score. While some instructors may dislike mandatory attendance policies because they can be a lot of work to enforce, there are recent technological advances such as “dibbers” (used at the Lancaster University geography department) or “clickers” (Hoekstra, 2008) which substantially ease the burden of collecting attendance data. Smart-card technology is available explicitly for the use of measuring student attendance; the TDS Student Attendance Monitoring Solution is currently being used at DCU. There are even new electronic systems which are being used to detect the ID cards students are carrying as they enter classrooms at Arizona University.

A backlash to all of this emerged this year in the Guardian newspaper, with the predictable question being posed: Is this new development necessary documentation or 'Orwellian' surveillance? One comment reads: "Why should the university care whether students attend a lecture or not? By the time a student reaches the university, they should be responsible for their own schedule and actions. Seems to me to be a technological solution to a non-problem." Analogies can be drawn here with the libertarian paternalism debates in public behavioural economics. Liam went through these recently on the blog. Keeping retention and graduation rates high is a major priority for many universities, but are the measures described above too extreme?

UCD Attendance Survey

A short while ago I posted about lecture attendance at Irish universities. Recent research by myself and others from Geary has shown that approximately 12% of Irish university students claim to attend all of their lectures. Overall, the mean-level of percentage lectures attended is 83% in Round 2 of the Irish Universities Study, and 84% in Round 3. This is a self-reported behaviour, and one that is subject to much comparison of anecdote amongst academic instructors. Furthermore, there is reasonable ground to suspect self-reported lecture attendance to be over-stated due to the phenomenon of social desirability bias. Social desirability bias is a term used to describe the tendency of respondents to reply in a manner that will be viewed favourably by others; see Bound et al. (2001) on Measurement Error in Surveys for a discussion (Handbook of Econometrics: Vol. 5).

While benchmarking against official data is difficult in the case of lecture attendance, a data-comparison can be made with a comprehensive attendance survey (measured by head-count) that was conducted at UCD during the academic year 2008/09. Under the guidance of Gabrielle Kelly (2010), students in an undergraduate Survey Sampling class carried out a survey to estimate the attendance rate at lectures in science modules in UCD. Only first-years in the UCD College of Engineering, Mathematical and Physical Sciences and the UCD College of Life Science were included (this was due to student drop-out in first-year being prevalent in these colleges). The overall attendance rate was 47.3% (+ 4.4%). However, there is also a (statistically significant) decrease in attendance rate as class size increases. The figure below (taken from the report on the UCD Attendance Survey; 2009) shows a plot of attendance rate vs. enrolment. The higher attendance rates between 70% and 80% occurred for smaller class sizes of between 30 and 160, whereas the lower rates of 15% to 35% occurred for larger class sizes of between 185 and 485.


It should be noted that there are some instances of attendance rates above 80% (in the UCD Attendance Survey), which is close to the mean-level of percentage lectures attended in Rounds 2 and 3 of the Irish University Study. The Irish Universities Study includes students from all courses but the UCD Attendance Survey only includes students enrolled in STEM (science, technology, engineering and maths) courses. These STEM courses have a greater amount of lectures to attend, which could possibly result in lower levels of attendance (compared to non-STEM courses). The above consideration would lead one to expect a higher level of percentage lectures attended in the Irish University Study, compared to the UCD Attendance Survey.

In relation to potentially unresolved concerns following from the comparison of anecdote amongst academic instructors, it should also be noted that students may be attending more of their lectures in the recession than they used to beforehand. University students in the UK study for two hours and 12 minutes more (per week) now than they did two years ago (in 2007), according to the Higher Education Policy Institute (2009). Given lower levels of labour demand in the part-time jobs market, there is certainly less opportunity for students to allocate their time to work (i.e. diminished opportunity-cost of study-time). In addition, the evolving crisis in the graduate labour market may motivate students to be more patient; and achieve higher academic standards (1 in 3 men under the age of 25 are currently unemployed in Ireland).

Naked Lecturing

In an article in yesterday's Irish Times, Ferdinand Von Prondzynski calls for lecturers to adopt "Naked Teaching". Don't get too excited, he means ditching powerpoint slides.

"The standard approach – 36 slides spelling out all the key points, with the presentation printed out for everyone in the room – increasingly represents bad practice, as it may actually inhibit the intellectual connection between the presenter, the topic and the audience, creating an automated process of very little value."

Lecture Attendance at Irish Universities

"If I’m Not Learning, Why Go?" This is the title of a blog-post from last year by Stephen Kinsella. Here's an excerpt: "One answer MIT surveyors found was when students didn’t feel they were learning, they didn’t go. This is certainly because the penalty to not going to a lecture is reduced by the presence of online learning materials like power point slides and handouts." Here's a link to a newsletter about the MIT Survey: Clay, T. and Breslow, L. (2006) “Why Students Don’t Attend Class“, MIT Faculty Newsletter, XVIII(4).

According to the MIT survey, students’ attitudes toward lectures vary widely, from "I never miss them" to "they’re worthless," with most responses falling somewhere in between. Most students reported they try to attend lectures, and usually do, missing them from time to time as the result of academic, extracurricular, or personal conflicts. "When asked to estimate what percentage of their lectures they attend, about two of every three respondents (67%) estimated that they attend at least 90%, three of every four (76%) that they attend at least 75%, and more than nine in 10 (93%) that they attend at least half."

Recent research by myself and others from Geary has shown that approximately 12% of Irish university students claim to attend all of their lectures. 32% of students claim to attend 90% or more of their lectures. 47% of students claim to attend 80% or more of their lectures. 57% of students claim to attend 70% or more of their lectures. Finally, 67% of students claim to attend at least half of their lectures. However, one must be cautious in drawing comparisons between the results from the MIT survey and the survey of Irish university students. The figures quoted above from the MIT survey are based on 47 responses by students in one subject. The figures quoted above from the Irish Universities Study are based on 2,867 responses across all subject areas. And of course, another (perennial) problem is social desirability bias: that is, students may say they attend their lectures because they think they should (say that).

According to Paul Latreille (H/T Stephen Kinsella), "falling lecture attendance among university students has become a major concern in the last few years. While the precise extent of the problem may vary, the phenomenon exists across subjects and has been documented in several countries (e.g. South Africa, Australia and the US, as well as the UK - see for example the references cited in Clearly-Holdforth, 2006). Various explanations have been advanced, including intrinsic factors (e.g. interest, motivation, learning styles and preferences), extrinsic factors (e.g. socio-economic considerations (such as the need to work), family commitments, assignment deadlines) and factors related to the lectures themselves (e.g. quality, value, interest)." Those interested in the role of technology in student learning will find an interesting discussion throughout Latreille's article. Laitrelle also covers the topic of "paternalistic" approaches to encouraging lecture attendance.

In 1993, David Romer published an article which ignited a lively debate about mandatory attendance policy: "Do Students Go to Class? Should They?" Daniel Marburger has investigated the impact of enforcing an attendance policy on absenteeism and student performance. The evidence he produces suggests that an enforced mandatory attendance policy significantly reduces absenteeism and improves exam performance. Alternatively, attendance at classes could also be potentially improved if students made a contribution toward the cost of their education. Finally, another possibility may simply be to communicate to students the importance of attending their lectures. Research papers by Schmidt (AER, 1983), Romer (JEP, 1993) and Durden and Ellis (AER, 1995) all show that lecture attendance is important for attaining high grades. Other studies from Irish settings also show the same: Maloney and Lally (1998), Kirby and McElroy (2003) and Purcell (2007).

Addendum: as is often the case in relation to other questions, establishing causality between lecture attendance and grades is difficult. Papers by Park and Kerr (1990), Bratti and Staffolani (2002) and Martins and Walker (2006) discuss this issue well. I am also reminded of the cartoon below.