This article is published in the September 2011 issue.

Interesting Times as CRA Approaches 40


Welcome to the 39th year of the Computing Research Association! And for our academic members, welcome to a new academic year. For those of you fortunate enough to not have been reading national news or watching your retirement portfolio, this last year has been quite the wild ride in Washington, DC, home to CRA World Headquarters.

There’s a lot going on at CRA so my purpose here is to update you on CRA’s efforts and activities on behalf of the computing research community. As always, I suggest that you monitor our two main information conduits: CRA Policy Blog  and the Computing Community Consortium blog. We can also be found at computingresearch on Facebook, @CRAtweets on Twitter and through our YouTube channel computingresearch.

The CRA Board’s most recent semi-annual meeting was held in San Francisco in mid-July with a focus on several topics of wide interest to our community. The summer meeting is typically the one in which we look towards the future, whereas the winter meeting often is focused more on the nuts and bolts of operation. This July, as I will describe in detail below, the Board looked in particular at two issues directly related to the health of the computing research community and two issues focused on how CRA can best support the community.

CRA’s main mission continues to be to ensure the health of the computing research enterprise (in North America specifically) so we are always thinking about the big issues. Computing is a relatively modern discipline, without overly entrenched traditions and cultures but also one in which traditions and cultures have developed very quickly due to the explosive growth of the field. Thus one very high-level question that was a focus of discussion at the July Board meeting is: Are the culture and traditions of computing research optimal for moving forward?

CRA and others have been exploring this question via a range of approaches for some time now — see “Computing Outside the Box”; Snowbird 2010 “Peer Review in Computing Research”; Snowbird 2008 “Paper and Proposal Reviews” Is the Process Flawed?” as well as a number of articles and posts by many other concerned researchers in, for example, CACM, on related topics. I take care to emphasize that computing research has been phenomenally successful with contributions and impacts that boggle the mind so exploring these issues comes not from a hand-wringing perspective, but from a “let’s make sure we’ve got it completely right” perspective.

There are a number of issues we could be addressing (your list may certainly vary): our emphasis on conference publishing; our large number of specialty conferences and concomitant lack of a computing-wide uber-conference; a concern that short-term gains are being emphasized to the detriment of long-range innovations in our research; a concern that funding panels and conference/journal reviewers are not doing the job we need them to do; a concern that our tenure and promotion guidelines discourage the kind of research that should be going on; and a concern that our community, given our importance, is not sufficiently engaged in the broader science policy space. Plans are afoot at CRA to develop a comprehensive approach to exploring issues related to the field as a whole—stay tuned for more information.

One growing piece of the computing research enterprise is postdoctoral research. Late last fall, CRA pulled together an ad hoc committee, led by Anita Jones, to explore the issue of how the contribution of postdocs is of maximal value to the community, which most definitely includes the postdocs themselves. CRA had already entered this conversation through the Computing Innovation Fellows (CIFellows) Project established by CCC and funded by NSF. The committee developed a white paper which was posted online (https://turing.cra.org/postdocs) and responses were solicited from the community.  A number of thoughtful comments were posted and a number of conversations were begun.

This is an important topic:  according to Taulbee data, there are currently as many postdoc positions open as for tenure-track faculty—an enormous change in just a few years. Is computing going the way of other sciences wherein multiple postdocs create a large holding pattern for researchers? This pattern can be highly discouraging and disruptive. And, given that postdocs are an increasing piece of the computing research community, what can we do to ensure that the postdoc experience is productive for all parties, but particularly for the postdoc? Please continue to provide your thoughts as we continue to explore the role and care of postdocs.

In what was obviously a packed Board meeting agenda, the Board also focused on CRA’s Taulbee Survey, which remains a pillar of CRA’s activities on behalf of the community. Taulbee is entering its 37th year of soliciting, and providing, information on the production and employment of computing researchers. Over the years, Taulbee has been expanded to include the production of computing students at all levels, salary and demographic data for faculty, and gender and ethnicity breakdowns. We are always aware there is much additional data that people would find useful, while simultaneously trying not to overburden our departments with even more extensive requests. And historical trends are of considerable interest so we are careful about making changes.

But it’s a good idea to periodically reevaluate what we do ask for and collect against what we should ask for and collect. This is particularly true this year because we have a “seismic” event to deal with. Recall that we have been publicly reporting Taulbee data in groupings based on the 1995 NRC rankings of PhD programs. We are all well aware that these rankings are only vaguely related to what people would consider reasonable rankings of programs in mid-2011, but they have had the virtue of consistency. And, since the NRC announced in the early 2000s that it would be providing revised and more current rankings, we have held off modifying our reporting scheme. But the NRC botched the job, both in terms of what data they collected, how they collected it,  and how they computed rankings from these data (see CRA Statement on NRC Ranking of Graduate Programs at www.cra.org) — and we simply cannot use the NRC results. Some folks suggested using the US News & Word Report data, but there are considerable issues with that approach as well.

CRA’s Taulbee Survey Committee, under the leadership of Stu Zweben (Ohio State), long-time Taulbee maven, tackled this issue. The first thought was to ask each survey participant to list its peers and to form reporting groups based on these self-selected peer groups. Even with the tools of our trade, this approach failed because the resulting groupings were not tightly bound. But, by looking at a variety of factors, the committee was able to develop a stable set of groups based on institution type (public or private) and faculty size — with an additional grouping based on the urbanization status for salary information. In order to assure a smooth transition, data from five prior years will be brought into this new reporting structure and provided as well.

The committee explored one more mechanism to provide additional value to our members: those departments participating in the survey will be able to provide CRA with a self-selected peer group and will receive a customized report showing their department in comparison to their peers. We are still working on the privacy and resource issues involved with this effort so stay tuned for more details.

The fourth area that the Board explored was the status and focus of our relatively new committee on education — CRA-E. When originally conceived, CRA-E was intended to provide a focus for computing researchers’ interests in, and activities in support of, undergraduate computing education. The original mission statement was quite broad, certainly too broad for an organization of (very) finite resources such as CRA, and it overlapped too much with activities already underway at our member societies. So the first activity of CRA-E, under the leadership of Andy van Dam (Brown), was a targeted effort, “Creating Environments for Computational Researcher Education,”  and a session at Snowbird 2010, “CRA-E Report on Basic Computing Knowledge”. Note the emphasis on “research”; this is what CRA-E can bring to the table.

CRA-E, with new leadership provided by Charles Isbell (Georgia Tech) and Ran  Libeskind-Hadas (Harvey Mudd College), will be continuing this approach by focusing on the health of the research pipeline in order to support interventions that improve the pipeline’s health. In particular, the idea is to focus on interventions that improve the quality of the pipeline rather than its numbers, and that are not already well covered by other efforts. As an example, CRA-E is seeking to answer the following kinds of questions: Where do our graduate students come from? Are there particularly successful institutions or programs that encourage the best and brightest students to pursue a career in computing research? If so, what can we learn from them?

In addition to these topic areas, there remain CRA’s core areas of ensuring that computing research is well-funded and well-supported by our society. It has an overwhelming impact on the quality and health of that society at all levels and ensures that the flow of highest-quality computing researchers continues

I’ll close with some staff notes. Patrick Krason has decided to leave the world of invoicing and member renewals for a position in a law firm dealing with compliance issues. Sandra Corbett has stepped into his shoes. And Carla Romero, CRA’s Director of Programs over the past nine years, has decided to return home to the sunny Southwest by the end of 2011. We are currently in the process of hiring a staff member to work with our committees to ensure that the many programs directed particularly towards diversity are well handled.

These are indeed interesting times. And CRA is both well placed and energized to work on the issues that will keep computing research and its people vital. So be sure to monitor, subscribe, like or follow us. Or good old-fashioned email to abernat@cra.org.

Interesting Times as CRA Approaches 40