An immodest proposal for conference organizing

Yesterday, my 15,000 closest colleagues and I received an email from the meeting staff of the American Educational Research Association (AERA), telling us that the schedule for the April meeting would be set six weeks before the meeting itself. The meeting is five days long, and attendees with presentations currently have no idea when they’ll be obligated to be in Toronto. Meanwhile, could we please register and set up hotel reservations?

We all submitted proposals by July 23, about eight and a half months before the conference start. We received acceptance notices November 1, more than three months after the submission deadline and more than five months before the conference itself.

The current system is only barely acceptable to those of us who are tenured faculty and who can afford to make reservations assuming we’ll be at AERA for all or most of the scheduled conference–i.e., it doesn’t hurt us too much if we have obligations on the very first and very last day of the conference, and have no clue right now, two months before the meeting, when we have to be present.

To untenured colleagues or graduate students without financial backing or reserves, the AERA annual meeting too often feels like a professionally-mandated event operated in a way to maximize anxiety and financial risk.

AERA staff and those who have been involved in program operations in the past will say quite truthfully that it is a large meeting that is complicated, and everyone puts forth a valiant effort to minimize all of the stress.

That is not enough. It is long past time for AERA meeting organization to change. We can no longer disrespect our less-secure colleagues and graduate students in this manner.

To put our experience in context, the American Geophysical Union (AGU) has a fall meeting that is roughly twice the size of AERA. For its December 10-14, 2018, meeting, the submission date for proposals was August 1, 2018. Somehow, AGU can manage the entire proposal-to-scheduling process with a submission deadline roughly four months before the start of the meeting, or half the time that AERA takes. Twice as large, half the time.

There are a few things that help AGU. First, they require a fee from scholars in industrialized countries for each submission. (They have a waiver process for those unable to afford the submission fee.) That fee allows AGU to pay professional staff to facilitate the reviewing and scheduling process. In contrast, AERA presidents have to recruit program co-chairs and beg them to run the meeting process. These program co-chairs are volunteers, and from my conversation with one of them a few years ago, AERA provides no funds to pay for support at the program co-chairs’ institutions. At each level of the organization (division, section, SIG), there are likewise volunteer annual-meeting organizers who run the reviewing/decision-making process, eking out time in late summer and early fall.

Second, AGU … actually, I have no clue about anything inside AGU’s conference process, as I am not a geophysicist. So I will not pretend to know anything else.

What I do know is that AERA’s process is crippled by the belief/culture that it can run a finely-tuned rating/ranking/sorting process with a largely volunteer program staff, with all results announced simultaneously and leaving no one behind, and have the result be anything other than constantly-late and putting the meeting calendar behind the eight ball… every single year.

This is nothing against the voluntary nature of AERA conference reviewing. I was a union activist for more than ten years in Florida. I was the education network organizer or co-organizer for most of a decade at the Social Science History Association. Yay for volunteers! But you can’t put the current structure on the shoulders of all the AERA volunteers and expect anything other than the disaster we’re seeing this year… and almost every year. To steal from the engineering dictum, you can have any two of the following: reasonable acceptance/rejection turnaround, volunteer organizing, or detailed ratings and mandated review feedback.

The solution? I can think of at least two: Give up either the voluntary organizing of the conference program (yay, fees) or dramatically simplify and shorten the reviewing process. I’d prefer to simplify the process in a few ways:

  1. Reduce what reviewers give feedback on. By default, reviewers today are asked to rate proposals on a 1-5 scale on a number of questions, and required to give feedback. I give huge kudos to AERA for wanting to structure the reviewing process so that everyone gets productive feedback. But that hasn’t helped shorten the process at all. My immodest proposal on this score: ask reviewers to give a simple yes/no answer to the following questions, with reviewers required to give feedback only if one of the answers is no:
    • Is the issue important?
    • Is the proposal professional?
    • Is this likely to be completed/ready for AERA?
    • Does the proposal fit this division/section/SIG?
  2. Change the acceptance process to one that is simpler and fairer: If a proposal gets the equivalent of three reviewers’ thumbs-up, it gets tossed in the Accepted pool. If the proposal gets two reviewers’ thumbs-up, the division/section/SIG program chair can decide to toss it in the Accepted pool or not. Then from the Accepted pool, pick items randomly until all the available slots for that division/section/SIG are filled. Yes — pick randomly from the acceptable proposals.
  3. Every division/section/SIG that completes its reviewing process can notify all submitters when the division/section/SIG is done. No waiting until the stragglers are done — this will make public which divisions/sections/SIGs are left and put pressure on (or better yet, get help to!) those program chairs to finish the process.

AERA doesn’t even need to try this all at once. It could allow divisions/sections/SIGs to volunteer to try this for the 2020 annual meeting… and let them participate in that third provision (early announcement of results!) if they do.

That won’t solve all the challenges of organizing the AERA meeting, such as reviewers who never respond to reviewing requests. But it can dramatically shorten the reviewing time, and thus lead to an earlier start on scheduling sessions.