Kathryn McKinley on double-blind reviewing

I recently had the great pleasure of meeting up with Kathryn McKinley, a faculty member at the University of Texas at Austin. She let me know that she is publishing an editorial pushing for double-blind reviewing in SIGPLAN conferences. The editorial is EXCELLENT. Here is a snippet:

“The purpose of double-blind reviewing is to focus the evaluation process on the quality of the submission by reducing human biases with respect to the authors’ reputation, gender, and institution, by not revealing those details.”

Simply put, sexism and nepotism are real. They may be unintended, but they happen. Double-blind reviewing improves quality and ensures that the best work really does rise to the top!

Comments about the Campaign Trail

This isn’t about computer science or academia, but I think it’s important to discuss other current events, such as the American presidential election. I have been disturbed by some of the sexist comments in the campaign. The comment that bothered me the most was this one:

“I understand that Senator Clinton, periodically when she’s feeling down, launches attacks as a way of trying to boost her appeal.”

Periodically when she’s feeling down? You have got to be kidding me.

“You challenge the status quo and suddenly the claws come out,” is another comment made by a presidential candidate.

There haven’t been any apologies made for these statements. I simply refuse to vote for someone who talks like this publicly in a campaign. I have four wonderful female students, a sister, a mother, two grandmothers, and a beautiful little 2-year-old niece, and when they don’t like something, it is not because they “feel down periodically” — it is because they have a voice and a right to share that voice.

Male vs Female Debugging Strategies

I had a nice conversation with Margaret Burnett yesterday. She is a faculty member at Oregon State University, and one of her research interests is end-user software engineering. She and many others — Neeraja Subrahmaniyan, Laura Beckwith, Valentina Grigoreanu, Susan Wiedenbeck, Vaishnavi Narayanan, Karin Bucht, Russell Drummond, and Xiaoli Fern — wrote a paper on male vs female debugging strategies published in CHI 2008. The subjects were not programmers, and they were debugging spreadsheets.

Roughly speaking, they found that successful men used dataflow (slicing) and testing strategies, and that successful women used code inspection and specification checking strategies. What a difference! Is it a difference between global thinking and sequential thinking, perhaps?

guest post by Jaymie Strecker: Graduate Employee Union at UMD

Here at parolefemmine, we don’t shy away from controversial topics. Here is a guest blog post by Jaymie Strecker, graduate student at UMD. Opinions, of course, are Jaymie’s and not necessarily mine.

Graduate Employee Union at UMD

By Jaymie Strecker, Graduate Student and Research Assistant,
Department of Computer Science, University of Maryland, College Park

Women in computing careers aren’t as numerous or as successful as we could be. To use Fred Brooks’s terminology, it’s not the essential complexity of computer science that holds us back—women can learn and apply computer science just as well as men—but the accidental complexity of the way the computing field has been implemented. Studies have identified a slew of difficulties stemming from this accidental complexity (this study is a good example), but I’m going to focus on two difficulties here: gender discrimination and trouble balancing career and personal life. I’ve picked these two because an organization is forming at the University of Maryland (UMD) that will work to resolve these difficulties for grad students—not just for women but for men, and not just in computer science but university-wide. The organization is
Maryland Teachers and Researchers (MTR), a union of grad employees.

You might think that gender discrimination is adequately addressed by UMD already. With a strong non-discrimination statement and an array of equity and diversity initiatives, UMD has striven admirably toward gender equality. Yet at least one gaping gender gap remains: the inexcusable lack of guaranteed maternity leave (let alone parental leave) for grad employees. The current policy places a grad employee seeking maternity leave at the mercy of her faculty employer—an uncomfortable situation at best, and an intolerable one if the faculty member refuses to grant the leave.

The subject of maternity leave leads into the difficulty of balancing career with personal life. To some extent, this may be an essential difficulty of being a grad student—a job which none of us expects to be easy or glamorous. But accidental complexity arising from our jobs as teaching assistants, research assistants, and administrative assistants—specifically, our vulnerability to exploitation by our employer—make the career-life balance even harder to achieve. A survey by UMD showed that teaching assistants were working more than the 20 hours per week they were paid for—not 1 or 2 hours more, but an average of 9 hours more per week. Grad assistants have never been paid much, and our pay raises are not even keeping pace with the cost of grad housing. With our low salary and the shortage of housing, it’s extremely difficult to find a safe, affordable place to live in the College Park area. All of these unnecessary difficulties can rob grad employees of time and options to pursue their personal lives.

How can MTR help? By giving grad employees the ability to participate in University decisions that affect us. By empowering grad employees to negotiate our work contract with UMD. By providing a meaningful grievance procedure for grad employees in case that contract is violated. Learn more at mtrumd.org.

MTR needs your help. Grad students, you can join at no cost. Faculty and staff, you can show your support (like Vibha has) by adding your name to the list of faculty and staff allies.

report from CHI 2008

I am finally getting around to talking about my great trip to CHI 2008 in Italy.

I had a great time!

The paper that was the most interesting for me personally was Investigating Statistical Machine Learning as a Tool for Software Development by Kayur Patel and others at UW’s DUB group. The paper talks about programmers who are using machine learning in their applications, but who aren’t experts in machine learning per se. The libraries they use are essentially black boxes to them, and that causes problems when you don’t understand the behavior of the ML library. For example, raw data is first converted into features and then is classified by the learner, and the mapping between raw data and classification result can be hard to trace. Another problem is difficulty understanding the tradeoff between time/space performance and accuracy. I thought this paper was interesting, and that in the future there will be more papers about people using complex libraries or components.

Lots of hallway talk at the conference was about the session called “Usability Evaluation Considered Harmful.” In this session, Saul Greenberg and Bill Buxton argued that good research was being rejected because it was too different or unique for a successful usability evaluation, and that boring research was being accepted because all the evaluation checkboxes were checked. I completely agree with this sentiment; I think the CHI community as a whole should reflect and change. In addition, I think that academic research needs to be as far out as possible. We need to do what companies couldn’t possibly conceive of spending resources on, and if that means it is too far out for a usability evaluation or just doesn’t have one, then so be it. I think good ideas will get out no matter what, and if CHI ignores them, then it will just become irrelevant. Hopefully that never happens and the CHI community changes course!

Vibha gets two nice letters from Larry

I know there are definitely some junior female faculty who have a lousy chair. But not me. My chair, Larry Davis, is great. He’s very nice to me, and he really supports me. He gave me space for my new HCIL UX Lab that I share with Francois Guimbretiere (UX stands for User Experience) and he gave me money to buy equipment too. He listens to me and I really appreciate his support.

Recently I got two letters from Larry. The letters include such tidbits as “If there is any way the department can help, do not hesitate to ask! Your high quality teaching is very much appreciated” and “I do want to thank you for your contributions that you have made to both the Department and UMIACS.”

I write this so that others can compare their Chair to Larry.

Now, I know that Larry has made some controversial comments in the past. In particular, when asked about recruiting women to computer science, I believe Larry said something to the effect that this problem begins very early, middle school or earlier, and it is difficult to try to address it at the college level. This is not a quote, just my recollection of what he said, so I may have gotten it wrong. I know that statement upset some people, but please note, he is right. It is a problem that starts very early. And also note the concrete ways in which he helps junior female faculty: me, and previously Lise Getoor who recently got tenure. I prefer people who act instead of people who say the right things but never do anything.

Harvard Business Review article on women in science

Great thanks to Angie Wu for letting me know about
this blog post
! It tells of an upcoming Harvard Business Review article that says that
“52% of highly qualified women working for science, engineering and technology companies voluntarily leave their jobs, driven out by hostile work environments and extreme job pressures.” (quoting the blog post here)

Since I don’t work for a company, I don’t feel qualified to respond or concur. It would be interesting to know the voluntary departure rates for women in academia. I don’t know what they are.

Double-blind review favours increased representation of female authors

Double-blind review favours increased representation of female authors
by Amber E. Budden, Tom Tregenza, Lonnie W. Aarssen, Julia Koricheva, Roosa Leimu and Christopher J. Lortie, in Trends in Ecology and Evolution, Volume 23, Issue 1, January 2008, Pages 4-6

Abstract: Double-blind peer review, in which neither author nor reviewer identity are revealed, is rarely practised in ecology or evolution journals. However, in 2001, double-blind review was introduced by the journal Behavioral Ecology. Following this policy change, there was a significant increase in female first-authored papers, a pattern not observed in a very similar journal that provides reviewers with author information. No negative effects could be identified, suggesting that double-blind review should be considered by other journals.

Whoa! When I heard about this paper, it was a confirmation of something I had already suspected.

My research area is software engineering and HCI. In HCI, there are double blind conferences, such as CHI. However, in software engineering, major conferences such as ICSE and FSE are not double blind. People complain that anonymization takes time. To me, the choice is very clear: do we want fairness and the best science published, or do we want implicit bias and a compromised review system? A little effort exerted by all seems more than worth it. Every conference should be double blind, in my opinion!

Women Don’t Ask

Linda Babcock, faculty member at Carnegie Mellon, wrote a famous book called “Women Don’t Ask.” As it says on Linda’s webpage, the book “describes her research on initiating negotiations and explores the societal factors that hold women back from asking for what they want.”

My friend Tanya Tarr recently was able to hear Linda Babcock give a talk to her group WIPT — Women in Politics and Technology. Tanya has great notes here — so if you don’t want to read the whole book, get the cheat sheet
here
!

It has been my experience that women don’t ask, and that experience is not just about salaries. It is about all kinds of professional opportunities in every dimension. Do women ask to teach what they want to teach, or to be added to grants and research projects? I’m sure some women do, but definitely this woman is guilty of not asking for nearly enough.

I remember that I was once offered an internship salary of $3000/month. My boss woke me up to give me the offer and I remember mumbling, “ok.” I later learned that male interns were making $4000/month. I didn’t ask what others were making, and so I didn’t know until it was too late. I didn’t ask for more, just to see if I could get it.

I’ll be honest. I think women are actively discouraged from asking for things in the workplace. The fault doesn’t solely lie with the woman, because she doesn’t ask for what she should. I feel that there is implicit pressure for women to be ACCOMMODATING.

I know of one female academic who is very successful and also isn’t afraid to get what she needs. I heard others complaining that she sends out way too much email. I don’t feel that way! The email she sends is the email she needs to send to get her job done. I suspect many men send out the same amount of email frankly. But she is perceived as “burdensome” for some reason.

Welcome to Parole Femmine!

Welcome to my blog. My name is Vibha Sazawal, and I am an assistant professor in Computer Science at the University of Maryland, College Park. I am also a wife, daughter, sister, aunt, and friend.

The Maryland state motto is “Fatti maschi, parole femmine.” Sometimes you see it as “Fatti maschii, parole femine,” but that is an archaic spelling. It is Italian for “actions are male, words are female.” Sometimes it is interpreted as “strong deeds, gentle words” — I personally believe that is the intent of the phrase.

Well, this blog is for “parole femmine” but I can’t promise that it will always be gentle.

I am recruiting tech and/or academic women to write guest posts on my blog. My hope is for this to be a place for women to share stories and express provocative views. I might also feature computer science research that I find particularly intriguing.

« Older entries