Thoughts on Lecture Capture

In thinking over whether I would consent to allow my lectures this year to be recorded, and the recordings made available to undergraduates, I came across this article by David Huyssen of the University of York. It points out that the papers most often cited in support of Lecture Capture are seriously flawed.

The technology that is offered at Oxford in the Department of Computer Science produces a sound recording of the lecturer's voice (provided they stay close to the lectern), together with images of slides that are projected from the Windows computer in the lecture theatre. I'm not sure if images will be recorded if the lecturer brings their own laptop and connects the data projector to that. There is no video recording of the lecturer themselves, nor any record of what is written on the whiteboard during the lecture.

Update 2022–3: The technology is now more sophisticated, though I doubt the automatic transcript can ever be made usable for science lectures.

My chief fear about lecture capture is that it changes the nature of the interaction, from being something that happens 'in this house, on this morning' to being something set in stone forever. It's not so much that I might say something I later have cause to regret, though of course explanations invented on the spot in response to discussion inevitably contain:

• Technical errors that do not interfere with the discussion on the spot, but would be confusing to anyone trying to pin down every detail by replaying the recording again and again. Such errors are easily corrected in the lecture if spotted there, but less easily fixed if they are part of a recording viewed long afterwards.
• Failed metaphors: improvised explanations sometimes fail to illuminate the subject. If that happens as part of a healthy, interactive lecture, then both lecturer and audience can shrug their shoulders and move on, perhaps to try again next time with a different metaphor. Having the failed attempt recorded and replayed later makes the mistake much more serious.
• Regrettable mis-speakings: ...

It's rather that caution to avoid these errors will be a real distraction from the task of giving the best lecture possible – here, today, in front of this audience. And if it's a distraction for me, an experienced lecturer, just think how the prospect of having their (perhaps naive) questions recorded will discourage a shy undergraduate from participating.

A briefing from OUSU

The Oxford University Students Union has published a polemical document, giving faculty reps guidance on how to campaign for Lecture Capture. It refers to a number of papers in support of the idea that "There is significant evidence that students will not be dissuaded from attending lectures by the existence of recorded lectures." Although numerical lecture attendance is not the only issue to discuss, it's worth pausing to consider it, as the only issue about which any evidence at all is offered in the briefing. The papers cited are as follows (quoted text in italics), and the first article cited tellingly contains bibliographical details for all the others.

• Karnad (2011) discovered that “students use lecture recordings to reinforce their understanding of lecture material, rather than ... as a replacement for attending lectures”.

This non-peer-reviewed article (local copy) is extensively discussed by David Huyssen. Since the article itself is a survey of other publications, it is meaningless to say that Karnad 'discovered' anything, and misleading to view it as in any way independent of the papers it cites.

• This is supported by Franklin et al. (2011),

Discussed by Huyssen. The paper is based on a survey completed by 206 medical students, and a comparison of student perceptions with actual exam results. Of those using the recordings, 14.4% admitted to a decline in their lecture attendance. Though a majority of respondents felt the recordings improved exam performance, the researchers concluded that "lecture recordings did not have an impact, either in a positive or negative direction, on exam performance across seven first- and second-year basic science courses."

• Soong et al (2006),

In a survey of 1106 students at Nanyang Technological University, two of the main reasons why students access video recorded lectures were given as, "I do not have to get up early for lectures," (7.5%) and, "I am too busy to attend classes," (4.5%).

• Traphagen et al (2009),

Also extensively discussed by Huyssen. It seems difficult to see how a paper can be cited in favour of this proposition when its results section begins with the sentence, "Class attendance counts were lower in the webcast section than in the non-webcast section for most lectures."

• Gosper et al. (2008)

One of the key findings of this report is that "Introducing [Web-Based Learning Technologies] will change lecture attendance patterns and may raise questions about the role of lectures," and the authors note that "Students appreciated the flexibility offered by WBLT with 75% using the technology because they couldn’t come to class and another 69% because it was the only class they had on that day."

• and Massingham et al (2006).

In a survey of 115 students in the Faculty of Commerce at the University of Wollongong, participants were invited to rate reasons for their non-attendance at lectures on a five-point scale. The reasons included statements such as, "I was genuinely sick", "Too busy", "Had to work", and "I can get the lectures on eduStream". Those whose attendance was rated as poor chose a significantly higher degree of assent to almost all reasons, including the availability of lectures online. It is difficult to see this as an unambiguous rejection of the hypothesis that lecture capture affects attendance.

Other links

Here are various UCU documents, most stemming from other institutions where less respect is (to be fair to Oxford) shown to the preferences and wishes of lecturers.

• https://roehampton.web.ucu.org.uk/campaigns/lecture-capture
• https://roehampton.web.ucu.org.uk/campaigns/lecture-capture/recording-lectures-faqs
  • "[T]he research on the benefits of lecture capture is at best equivocal. Reputable research (there is much that is shoddy and / or published by firms with a financial interest in lecture capture software) suggests that attendance falls when recordings are made available. Marks often decline as well. It looks as if high-performing students who would get top grades anyway, benefit slightly; poorer students, those who are working or have home situations not conducive to study do significantly worse. Access to recordings reduces note-taking skills and evaluation of material, encouraging surface learning."