High-level skills, low-level tasks (#1252)

Topics/tags: Miscellaneous, technology

Today, I’ve been working on various tasks for the SIGCSE Technical Symposium on Computer Science Education. Many of the times that I work on these kinds of tasks, my lovely wife [1] says things like

You’re a professional. You should be using your high-level skills, designing policies and such. Why are you doing low-level tasks like answering emails and figuring out why people can’t sign in? Is that really the best use of y our time? Shouldn’t someone else do that?

It’s a reasonable question, one that I should ask myself more regularly. I’m one of the Program Chairs for SIGCSE because I have a lot of experience in CS Education, because I can think about issues like our reviewing process or how to balance the program, because I can work with others on complex processes [2].

However, as Program Chair, I also get tasked with many low-level tasks. What kind of tasks? Let’s consider a few that I or my co-chairs worked on today.

• We helped two of our reviewers figure out why they couldn’t log in to EasyChair, our submission and review software [3].
• We explained to someone why they could not submit their work after the deadline.
• We moved some papers from one track to another.
• We reviewed about thirty papers to see if they met the formatting and content guidelines.
• We changed the email address for one of our reviewers.
• We answered questions about the formatting guidelines.
• We handled messages from reviewers who volunteered to review more papers [4]. That often included responding with a thank you and updating EasyChair to add the new role.

Some of these are high-level tasks, such as explaining our policies to someone or thanking our reviewers. I suppose we could leave those to generative AI, but that seems inappropriate.

On the other hand, most of these are low-level tasks, particularly working with the software. It seems like we could hand them to a graduate student or even an undergraduate. So why don’t we? There are a variety of issues at play. Let’s see …

• Much of the information we deal with, such as author and reviewer names, is confidential. It doesn’t seem reasonable to give others access to the information. We don’t even share it with members of the Program Committee. I don’t think one can do the small tasks without having access to the information.
• There’s no budget for it. It might seem like a moderately large conference like the SIGCSE TS [5] would have a lot of money to work with, but it doesn’t. In most years of late, it has only come close to breaking even [6].
• It’s hard to find people trained in using EasyChair. And, as I’ve noted before, it’s not the most intuitive software. While we could train folks, I’ve found that we have enough experience that we are likely better at figuring out how to do things.
• Some of these tasks seem like it would take about as much time to tell someone to do them as it would to do them. I suppose we could reverse the process; our assistants could process the email and only pass the high-level tasks to the chairs.
• The work is intermittent. I spent five to ten minutes each hour rather than doing the work in one block of time. Is that important? Yes, I think so; people feel better if they get prompt responses. And one of my implicit jobs as Program Chair is to make people feel valued and included [7]. It’s also intermittent in another way; some weeks we have almost no work. Saturday, I worked for seven hours.

Hmmm … the first and the last are the big sticking points. And I don’t know how to deal with them. I’ll also admit that I don’t like giving up control of delegating work; I’m pretty sure my co-chairs feel similarly. Even if we could deal with our reluctance to delegate, privacy remains a significant issue.

And so I’m left doing low-level tasks in my high-level position. C’est la vie. At least I sometimes I have fun with the low-level tasks. I’m strange that way.

I wonder what other conference program chairs do.

Postscript: Yes, Michelle, you’ve convinced me that I should delegate more. I’m just unsure how to do so here

[1] Also my only wife.

[2] Yes, really.

[3] I like to say that EasyChair is like FastLane; at the time it was written, it was likely groundbreaking software. But that was twenty years ago. Twenty years later, the UI seems clunky, at best.

[4] I so appreciate that the SIGCSE community is coming together to help us deal with the large number of submissions.

[5] The technical symposium generally attracts about 1500 attendees. I’m told that makes it one of the five largest ACM conferences.

[6] My lovely wife would also tell me that we could still budget for assistants.

[7] They are. That is, we value our reviewers and our authors. We wouldn’t be a community without them.

Version 1.0 of 2023-08-14.