SAN FRANCISCO, CALIFORNIA—At most scientific conferences, almost every nametag you see dangling from people's necks shows a university title. But this wasn't the case for many people wandering the halls last month at the annual meeting of the Association for Psychological Science (APS). Many of those tags featured names like Google, Microsoft, Facebook, and other tech giants. It’s a sign of the times: Social scientists are getting snatched up by tech companies.
Mary Czerwinski is a cognitive psychologist based at Microsoft Research (MSR) in Redmond, Washington. Czerwinski earned a Ph.D. in cognitive psychology at Indiana University, Bloomington, and then she jumped immediately to industry, doing human-computer interface research for a series of companies while holding a part-time teaching position at Rice University in Houston, Texas. She made the jump to Microsoft 18 years ago, first working on products for the software company and then moving, after 2 years, to MSR. Czerwinski has done everything, from software interface development to basic research on attention. Today she leads a team at MSR working on several projects, including one on clothing that senses its wearer’s emotions. We sat down with Czerwinski between sessions at the APS conference and asked her about her work. This interview has been edited for clarity and brevity.
If you can think like a programmer, you're far more efficient at working with them on projects.
Q: What's it like being a social scientist in a tech company rather than a university?
M.C.: I don't have to write grants! But you do have to work in a team.
Q: What does the team do?
M.C.: I manage a team that researches the future of human-computer interaction. Nowadays, we think more about social computer interaction; it’s not just one person in front of a computer anymore. For the most part, we are logging and sensing what you do and how you react to computer/social interactions and anticipating your needs based on your context and past behaviors.
Q: What's the goal?
M.C.: We're trying to make your interactions with computers more pleasant, positive, and productive. You need a computer system that is more emotionally sentient but that also remembers how you prefer to interact with a system. Others in my group try to visualize data in a way that is easier to use and interpret so that you can understand patterns and trends in data that are relevant to you.
Q: Many social scientists worry that getting a job at a tech company pays well, but then you spend all your time tricking consumers into clicking on advertisements, buying things, or giving away private information. Do you get to do any basic research?
M.C.: I actually started in product development at Microsoft, managing a usability group. We did epic things there, like play-testing games, Barney the robotic doll, Expedia, CitySearch, Microsoft Money, and Encarta. After 2 years, I was wooed by researchers to join MSR, and it was the best move I ever made. But even before I moved to MSR, I was able to do basic psychological research, even when I was working on product teams. It’s all about how you frame your study questions. Sure, if you let a product team pressure you to focus on a “feature” that they’re interested in, then you aren’t going to do productive research that furthers the team’s knowledge and efforts. When I do basic research, we ask questions, for example, about the benefits of doing visual search in 3D using particular kinds of distortion techniques or using specially designed input devices, and compare those innovations to existing practice. In those research studies, there are no products in mind, just the human-computer interface research questions. Applied research is always progressing toward a product you can see shipping someday. In basic research, you just have the scientific curiosity.
Q: You were the very first social scientist at MSR. How many are there today?
M.C.: My guess is that we have around 100 social scientists in MSR alone these days, worldwide, and I might be very shy with that number.
Q: How did you make the transition from academia?
M.C.: I went into industry right out of my Ph.D., first working for Bell Communications Research. Then I worked at Johnson Space Center, and I just applied to Compaq Computer on a whim—they were looking for human-computer interaction researchers. And I got the job! In 1995, I was an adjunct professor at Rice University while I worked at Compaq. They were amazingly generous in giving me my Fridays off from noon on, to go to psych department lectures, teach a grad class, and then attend Rice's visiting guest lectures at 3:30 p.m. I did that every other semester, and it saved my academic life. The first time I taught the class, my daughter was 6 months old. I remember staying up until three in the morning reading the material to prep for the class. It was crazy, but it kept me relevant to the field I had trained in. Then I headed to Microsoft.
Q: Is it tough being a woman in tech?
M.C.: I always tell young female researchers that, back in the day, the best thing I had going for myself in industry was the last three digits in my name: “Ph.D.” It did work. But I have to say that I have been amazingly lucky to have not suffered from gender bias for much of my career, even in graduate school. I have been surrounded by unbelievably supportive mentors, advisers, and peers. The one time I did suffer from gender bias, I consulted with my manager and then took it to him face to face.
Q: How much computer coding do you have to do as a social scientist in a tech company?
M.C.: By the time I got to Microsoft Research, there was no reason for me to be prototyping or programming anything. I was working with the world’s best.
Q: Should social scientists not bother learning any coding then? Is it not helpful?
M.C.: No, it's helpful. At the most basic level, it helps you talk to engineers. And learning a computer programming language fundamentally changes how you think, more than almost anything else. If you can think like a programmer, you're far more efficient at working with them on projects. And ultimately, if you have some coding skills, then you can fall back on that and do prototyping of projects yourself if you need to.
Q: What coding language do you find useful these days?
M.C.: Python! It's an attractive and user-friendly language.
Q: What do you miss in academia?
M.C.: I get the best of both worlds. I am an adjunct professor at the University of Washington and can be on many Ph.D. dissertation committees and advise students. I have students for interns, and I don’t have to write grants. I guess the biggest downside might be that I don’t get to be with those students for all of their 4 or 5 years. And I don’t get to teach day in and day out, which I do love. But the students do still check in with me all along their career paths, so I am lucky in that way.
Q: Any regrets?
M.C.: What has changed is that now there is really big data—we can log, file, track, and check so many things that we couldn’t before. I would have loved to have had a degree in social psychology, not cognitive psychology, but never look back!
A Random Selection of Czerwinski’s Publications
Laura Pina, Asta Roseway, Mary Czerwinski, and Kael Rowan, “In Situ Cues for ADHD Parenting Strategies Using Mobile Technology,” IEEE Pervasive Health, May 2014
Pablo Paredes, Ran Gilad-Bachrach, Asta Roseway, Mary Czerwinski, and Kael Rowan, “PopTherapy: Coping with Stress through Pop-Culture,” IEEE Pervasive Health, May 2014
Gloria Mark, Shamsi T Iqbal, Mary Czerwinski, and Paul John, “Bored Tuesdays and Focused Afternoons: The Rhythm of Attention and Online Activity in the Workplace,” ACM Conference on Computer-Human Interaction, April 2014
D. MacLean, Asta Roseway, and Mary Czerwinski, “MoodWings: A Wearable Biofeedback Device for Real-Time Stress Intervention,” PETRA (Pervasive Technologies Related to Assistive Environments) 2013, May 2013
Bongshin Lee, Nathalie henry riche, and Mary Czerwinski, “Understanding the verbal language and structure of end-user descriptions of data visualizations,” ACM Conference on Computer-Human Interaction, May 2012
B. Byun, A. Awasthi, P. A. Chou, A. Kapoor, B. Lee, and M. Czerwinski, “Honest Signals in Video Conferencing,” in Int'l Conf. on Multimedia and Expo (ICME), IEEE, July 2011
Margaret M. Burnett, Scott D. Fleming, Shamsi T. Iqbal, Gina Venolia, Vidya Rajaram, Umer Farooq, Valentina Grigoreanu, and Mary Czerwinski, “Gender differences and programming environments across programming populations.,” ACM ESEM , September 2010
Kamal Jain, Yang Song, Li-wei He, and Mary Czerwinski, “Evaluating the Unaccounted Cost of Distraction of Display Ads to the Users,” in Web Science Conference 2010 (WebSci 2010), Association for Computing Machinery, Inc., 2010
Kori Inkpen, Steven Whittaker, Mary Czerwinski, James Wallace, and Roland Fernandez, “GroupBanter: Supporting Serendipitous Group Conversations with IM,” in The 4th International Conference on Collaborative Computing: Networking, Applications and Worksharing (CollaborateCom 2008), Orlando, FL, November 2008
Mary Czerwinski, Greg Smith, Tim Regan, Brian Meyers, George Robertson, and Gary Starkweather, “Toward characterizing the productivity benefits of very large displays,” in (2003) Interact 2003, IOS Press, 2003
Mary Czerwinski, Edward Cutrell, and Eric Horvitz, “Instant Messaging and Interruption: Influence of Task Type on Performance,” pp. 356-361, Association for Computing Machinery, Inc., December 2000