Days and weeks often go by in a blur, packed with meetings and emails, leaving my “to do” lists perpetually unfinished. Where does the time go? Why is there so little time for learning new things, reading papers, working in the lab, and other activities that one might naively think would be part of a Physics professor’s job?
A while ago, I decided to log my time, keeping track of what I’m doing in 15 minute intervals during the day, and more loosely in the mornings and evenings. (The idea was inspired by my brother’s time as a corporate lawyer, when he had to account for his time in 6 minute intervals.) I’ll explain below why I found this a surprisingly useful exercise. First, though, I’ll list the categories and the results. I created nine categories:
- Emails
- Seminars, reading (research related), colleagues
- Analysis / thinking / writing
- Communication (talks, grants)
- Students (research), or lab work
- Teaching & course prep
- Service / administration / committees
- Planning / organization
- Other (walking, staring off into space, eating lunch, …)
Some activities could fit into multiple categories, in which case picked the one that seemed most appropriate. For example, reading and writing a set of emails relating to some committee would fall under Service (#7) rather than Emails (#1), while spending a block of time whittling away at all sorts of emails in my Inbox would fall under Emails. It’s not the ideal categorization, but it’s ok.
I didn’t keep this log consistently, but rather over 140 days between January 2016 and October 2017, usually in complete week-long blocks. I didn’t keep notes during summer, when categories 6 and 7 would be sparsely filled, and 3-5 more densely occupied.
What have I found? Here are the fractions of time spent on each category:
At first glance, this looks great — nearly a majority of my time goes to research-related activities such as seminars, reading, data analysis, and working with undergrads and graduate students in my group, and the next largest category is teaching. The overall distribution, however obscures some gloomier conclusions.
First, if I split the log into “8:30am – 5:30pm” and other times (basically, about an hour in the morning plus a few hours every evening), there’s an obvious difference: the large chunk of time spent on emails:
Emails take up a huge amount of time, and the only way to keep them from devouring my day is to slog through them in the evenings.
Second, the fraction of time devoted to “Service” (committees, etc.) has been growing:
That’s what I would have guessed, but with the numbers I have some reassurance that I’m not just perceiving time move more slowly in a constant number of meetings. Even though some of these activities are important, and even intellectually enjoyable (such as faculty search committees), many are not. I’m on a wide range of committees, ranging from our department’s graduate studies committee (of which I’m the chair, having been foolish enough to suggest ways of improving our program), to the creation committee for a new biological imaging core research facility (of which I’m also the chair), to the building committee for our new $500m applied / translational biosciences campus (of which I’m thankfully not the chair, though I expect a multi-hour discussion of chair design to come soon). There are other committees, too, related for example to teaching and outreach.
The “service” category also undercounts the time spent on service-related activities — some of this falls into the “email” bin, for example.
What have I learned from this?
First, I found the exercise of keeping a time log useful for making me realize more clearly how much various activities actually take. Often it feels as if writing an email “should” take a few minutes, when in actually it takes fifteen, which I would not have acknowledged without being forced to. Similarly, I have a better appreciation of how a 1 hour meeting invariably sucks up a greater amount of time, if one accounts for the “extra” time every required for walking across campus, scheduling, etc. It’s easy to ignore these things, and quantifying them makes their cost more apparent. (I recommend the activity, at least for a few weeks!)
Second, it provides a perspective on how the university might accomplish certain aims. We recently received, for example, a query from our Vice President for Research asking for input on things that would increase the research activity of the campus, such as providing a few thousand dollars for conferences to seed large grant proposals. In my opinion, a more effective approach would be to remove the service burdens on active faculty (and move them to inactive faculty). There are obvious problems with this, but it would be interesting and possibly useful to at least discuss paths like this.
Knowing where my time goes, can I better allocate it to things I actually want to do, like learning new skills, working on new data analysis methods, and spending more time with students in my group (especially actually doing things in the lab, which I hardly do at all at present)? I would love to study the machine learning methods my group is exploring (for an early foray, see here), assess different methods we’re implementing for measuring diffusion, and finally push the buttons on an automated microscope we’re building, but all of these take time. Plus, there are things I’d like to write and publish about courses I’ve developed, and there shouldn’t be a 2 month gap between my blog posts. All these types of activities aren’t frivolous: they’re a big part of the core mission of a research university. The answer: I don’t know. It’s not physically possible for me to spend more hours working, so something else has to be culled. I now know how much; the tough question is how.
Today’s illustration…
Appropriately, an unfinished painting of deer at a pond at sunset. Someday I’ll complete it…
The Eighteenth Elephant 