This is a whimsical topic that began with a Christmas present from my daughter, a copy of The essential Leunig: Cartoons from a winding path . Near the end of his truly beautiful introduction to the book, Michael Leunig wrote:
My cartoons have been described as whimsical, an idea I used to resist, but which now, having gradually strengthened my faith in the innocent genius of our whimsy, I would like to embrace. Proceeding along a Leunig-like, whimsical winding path, The Phrase Finder website gives an explanation for the origins of the phrase "Burning the midnight oil" that may resonate with academic researchers :
Meaning: To work late into the night. Originally this was by the light of an oil lamp or candle. More recently, the phrase is used figuratively, alluding back its use before electric lighting.Could this suggest to you, as it does to me in a somewhat whimsical  way, a vision of academics spending "the day in toyle", that is lectures, tutorials, committee meetings, student contact, admin, marking assignments, etc., then going on to "tyre the night in thought", that is, creative writing up one's research, for presentations to conferences or submissions to journals? An academic quoted by Tynan et al. (2012) , in the context of an interview question about hours spent in online teaching, described this type of time division in a more prosaic way:
Origin: The English author Francis Quarles wrote in Emblemes, 1635:Wee spend our mid-day sweat, or mid-night oyle;
Wee tyre the night in thought; the day in toyle.
Well, I think most people would spend their day at work mostly doing teaching and administration. And I think a lot of colleagues would have their nights and weekends and annual holidays as research. Going into free fall along the whimsical path, can the technologies help to explore whether this (perhaps romanticised) vision has some measurable or tangible connection to the actual behaviour of academics? That brings me to the topic of email time and date stamps for messages that submitted files for journal and conference publication. Having accumulated a rather large number of these over the years, I have a readily available data set for exploring this question. Of course there are many other technologically-generated data sets that could offer insights into actual behaviour. For example, Wang et al. (2012)  used records of download time stamps for articles in science journals published by Springer  to conclude that "many scientists are still engaged in their research after working hours every day" . However, a comprehensive review of other kinds of data sets and methodologies that may be used to explore the actual behaviour of academics is beyond the scope of this meander!
Figure 1 summarises an analysis of time and date stamps for submissions of "full paper" and "abstract only" proposals to Teaching and Learning Forum  during the years 2005-2013, classifying the submissions into "Office hours" (0830-1730, weekdays) and "After hours" (times other than "office hours").
Figure 1: Time and date stamps for email submissions for TL Forum publication, 2005-2013
The time and date stamp for a submission marks only the concluding time for a "work" (although in many cases authors sent pre-submission or follow up emails). The stamp does not indicate the distribution in time of the "work". For example, an author may have commenced an abstract during a weekday afternoon, posting it at 1731 hours (i.e. "after hours"); or an author may have burned the midnight oil, then posted it next morning at 0830 hours (i.e. "office hours"). Nevertheless, Figure 1 suggests a noteworthy amount of the time academics spent on TL Forum submissions was "after hours" time. Also, Figure 1 suggests a long standing (9 years) and stable pattern, given that during 2005-2013 there were only minor changes in TL Forum's call for papers (for CFP details, see ).
- Times are as written by the sender's email software (e.g. "18:32:41 +0800"), corrected where necessary to the sender's local time zone using a knowledge of their address (e.g. "02:56:48 +0000" - convert to "10:56:48 +0800"). The data was copied from emails, into the "all receivals" file recorded for each Forum.
- Times for TLF emails other than submissions were not included. The impact of omitting non-submission emails (requests for extensions, requests concerning scheduling, etc.) was checked for TLF 2013, for which all emails concerning submissions totalled 272 (25.4% after hours) compared with actual submissions totalling 110 (27.3% after hours). As this difference seemed small, counts were confined to actual submissions, as the "all receivals" files were easier to examine than the raw email archives.
- The number of submissions averaged 94 per Forum, range 73 (2007) to 110 (2013), with no discernable trends during 2005-2013.
Although a "noteworthy" amount of the time spent on TL Forum submissions was "after hours" time, Figure 1 can show no more than very modest support for "mid-night oyle" hypotheses. Some further evidence can be gleaned by the category of publication sought, there being two categories for TLF, "full paper" or "abstract only" . Full paper submissions during 2005-2013 totalled 272 (34.8% after hours) compared with all submissions totalling 845 (25.2% after hours) - a slight suggestion that the larger amount of work in a 5000-6000 word paper led to a higher incidence of "after hours", compared with the much smaller amount of work in a 200 word abstract.
Whilst this brief study is not conclusive, we need to persist in exploring the contribution that technologically-generated data can make to research into academic workloads and the nature and timing of academic work. Quite obviously, though it's wishful thinking, the process of receiving conference and journal submissions, and related communications such as reviewers' reports, could include the administration of an immediate follow up questionnaire about an "office hours" and "after hours" breakdown, that can be linked closely to non-subjective, automatically recorded data, as in time and date stamps. Also quite obviously, there are numerous other data sets that could be subject to automated "data mining" or "analytics", to produce academic workload and time distribution information. Some data sets are created and controlled by universities, for example time and date stamps for emails processed by university servers, or for learning management system actions, such as uploading a new assignment, moderating a student discussion forum, etc. Some data sets are created and controlled by publishers, for example data derived from journal submission processes, or from readers downloading articles, etc. (an example is given by Wang et al., 2012 ). Indeed there may even be the prospect for needing a caution about data mining, similar to a caution given recently by National Tertiary Education Union national president, Dr Jeannie Rea. In an article in The Australian newspaper , she was quoted (I have italicised the key phrases):
... Constant accountability has turned into a feeling of constant surveillance and the message is you need to be scrutinised; that there is something wrong ... All the evidence -- graduate outcomes, student satisfaction, international rankings -- would suggest otherwise, but the impression is academics need to be watched. As to the significance of academic workload problems, or "work/life balance" problems , or the problem of "hyperprofessionality" , here follows a few recent examples from the extensive literature. In discussing survey respondents' comments on academic workloads, Bexley, James and Arkoudis (2011)  included the following in a selection of typical comments:
The job hours are on average 70 hours a week and I usually work one or both days each weekend. At the moment it is Saturday night at around 7pm and I am fully engaged in work, about to read one of my PhD students' theses, and this will go on until at least midnight. A very typical weekend. Reviewing some literature on academic working hours, Gill (2009)  wrote that:
... academics and teachers were more likely than any other occupational group to do unpaid overtime. A large proportion were working hours in excess of the European Working Time Directive, and 42% said that they regularly worked evenings and weekends in order to cope with the demands of their job. Currie and Eveline (2011)  were quite explicit in their findings from a "work/life balance" study:
Most [respondents] felt that having e-technologies at home was of benefit to their work but they came at a cost to their family life -- delivering a blessing and a curse. A blessing and a curse? Well, there is one final detail to add. In accord with the "mid-night oyle" theme, this contribution was emailed to HERDSA News Editor Roger Landbeck at about "01:20:00 +0800" :-)
|Author: Roger Atkinson retired from Murdoch University's Teaching and Learning Centre in June 2001. His current activities include honorary work on the TL Forum conference series, Issues in Educational Research, and other academic conference support and publishing activities. In mid-2012 he retired from a 17 year association with the publishing of AJET. Website (including this article in html format): http://www.roger-atkinson.id.au/
Please cite as: Atkinson, R. J. (2013). "Burning the midnight oil": Time and date stamps on email for a conference publishing activity. HERDSA News, 35(1). http://www.roger-atkinson.id.au/pubs/herdsa-news/35-1.html