On 22 April 2024, the Office for Intercollegiate Services (OIS) published the first report summarising the results of the Undergraduate Supervisor Workload Survey conducted in January-February 2024. The report, prepared by OIS, is available here. Below, we summarise what we consider to be the key findings of the report.

OIS and the Colleges have produced an independent interpretation of the report here. Cambridge UCU and the Justice for College Supervisors Campaign (J4CS) disagree with certain choices of emphasis and uses of summary statistics in this document, which we explain below.

Key Points

•  The survey data is robust, with a large and representative sample of respondents.
•  A clear majority of supervisors are paid the intercollegiate re-charge rate for their teaching.
• Across all supervisors, the median time it takes to prepare for a one-hour supervision is between 2 hours 40 minutes and 4 hours 30 minutes.
• When preparation time is taken into account, a significant proportion of supervisors are paid less than the national living wage (£11.44 per hour) for their work under the current system.

Background

The survey was written jointly by OIS and Cambridge UCU’s J4CS Campaign, with input from the Colleges and the University, and administered by OIS. It was sent to everyone who had supervised between October 2022 and December 2023. The purpose of the survey was to establish a broad, up-to-date view of the population of supervisors and their labour conditions, including ascertaining who takes part in the supervision system, how much time they spend supervising, and how much they are being paid.

The survey also provided other basic data about the supervision system—for example, how common are different sizes of supervision group? How common are the different styles of supervision, e.g. essay- vs. question sheet-based?—which are expected to inform the ongoing University Teaching Review.

The published report summarises the responses to the quantitative questions asked in the survey. The analysis presented in the report was carried out by OIS and checked by a J4CS campaign member, who was not given direct access to the data. Analysis of the qualitative and free-response questions (e.g. “What motivates you to supervise undergraduates?”) is ongoing, and the results will be published in a later report.

The most recent data comparable to this survey are the results of Cambridge UCU’s 2018 survey of hourly-paid staff at the University, which addressed some of the same questions and received 140 responses. The summary report of the 2018 CUCU survey is available here.

Survey response rate

The response rate to the survey was very high: 1878 people, or 35% of recipients, completed at least part of it. This represents “38.6% of the average number of supervisors in an academic year” (Report paragraph 9, page 4). Respondents spanned career stages (Figure 2) and academic subjects (Figure 1). This is a more than tenfold improvement on the number of responses to Cambridge UCU’s 2018 survey, with the caveat that this new survey is very specifically tailored to supervisors, rather than all hourly-paid teaching staff.

Payment rates

73% of survey respondents earn the standard payment rate, known as the intercollegiate re-charge rate, for supervising (see table 9, page 12), at least some of the time. 67% of survey respondents earn this rate for all of their supervising. This highlights that the re-charge rate acts as a default supervision payment rate, even though some supervisors earn more than this based on arrangements with their Colleges (see tables 11-13, pages 12-13, for details).

Supervisors in early stages of their careers—PhD students, postdocs, and research fellows–are more likely to earn the standard rate than University Teaching Officers (UTOs), College Teaching Officers (CTOs), Directors of Studies (DoSs), and College fellows (see table 10).

Preparation time

The median amount of total preparation time per supervision, across all respondents, was 2.67-4.5 hours (see table 62, page 47). This accounts for marking student work, reviewing course materials, emailing students and DoSs, booking rooms, and other miscellaneous preparatory work, but does not include writing and submitting CamCORS reports (as these are done per student, rather than per supervision).

We quote this result as a range, because respondents were asked to estimate each component of preparation time as a range (e.g., survey question 10, “Typically, how much time does it take for you to mark the work of one student?: 0-10 minutes, 10-20 minutes, …”). This result is consistent with the results of the 2018 Cambridge UCU hourly paid teaching survey (see CUCU 2018 HPT report, page 14), where the median respondent reported spending 3 hours preparing for 1 hour of teaching.

Figure 6 (page 47) displays the distribution of total prep time per supervision across all respondents who answered the preparation time questions. Figure 7 (page 48) shows a more detailed view of the data, without any rounding or binning of responses, but cutting off at a maximum of 10 hours of prep time per supervision.

As these figures show, the distribution of total preparation time across all respondents skews rightward, i.e. it rises sharply and then falls gradually with increasing prep time. As a result, the average, or mean, amount of preparation time per supervision quoted in table 62, 3.5-5.23 hours, is higher than the median, and we feel that the median is the fairest statistic to use to summarise the data. Table 62 also quotes the mode of the distribution, or the most common response, at 1-1.67 hours. However, as figure 7 shows, this is the most common response (with 34 respondents) by only one person. The second-most-common response (with 33 respondents) is about an hour higher. For this distribution, the mode is not representative of the “middle”. As a result, we are disappointed to see the mode as the only summary statistic about preparation time quoted in the OIS interpretation document (point 6f).

As an indicative calculation, if a supervisor teaching a group of two students and earning the standard intercollegiate re-charge rate (£20.97 per student per hour, or £41.94 per supervision for a group of two) spent 2.67 hours on preparation and 1 hour teaching the supervision, then that supervisor would be earning an effective hourly rate of £11.43. This is just below the national living wage as of 1 April 2024, £11.44/hour.

Figure 8 (page 49) shows that there is a weak trend between years of experience supervising and prep time: more experienced supervisors spend slightly less time preparing. However, a substantial fraction of even the most experienced group of respondents (more than 5 years of experience) still spends hours preparing for one supervision: the median amount of preparation time reported b this group is 2 hours and 15 minutes.

Figure 12 (page 52) shows that the larger the supervision group, the longer the supervisor spends preparing. Relatedly, Table 22 (page 19) shows that most supervisors take between 30 minutes and 1 hour to mark the work of one student; again, there is a weak trend with years of experience supervising, where supervisors with more experience are slightly faster at marking. This is important because it implies there is a limit to the efficiencies of scale introduced by teaching students in larger supervision groups: more students will inevitably mean more preparation time spent on marking, regardless of supervisor experience.

Conclusion

The survey has provided Cambridge UCU, the Colleges, and the University with an extremely valuable and comprehensive view of the supervision system, far exceeding what was available before. This will be useful in ongoing discussions between J4CS and the Colleges related to the programme of work on undergraduate supervisions. Cambridge UCU and the J4CS campaign thank Malcolm Millbrook and OIS for the work preparing the survey report.

Members with questions about the survey and its analysis may contact the branch (admin@ucu.cam.ac.uk).