CHS: Small: Software to Support Diverse Problem Solvers

Oregon State University, Corvallis OR

Investigators

Abstract

Even though females are using computers and the Internet at rates similar to males, they hold only 26% of IT professional positions. Teenage girls are five times less likely than males to consider a technology-related future, and the number of first-year undergraduate women interested in majoring in computer science declined by 64% between 2000 and 2012. Alarmed by these statistics, the PI notes that evidence suggests many common software features were inadvertently designed around the way males tend to work, and actually undermine many females' problem-solving because of the many differences in the ways females and males tend to approach and solve problems, including for example their motivations for using software, their information processing styles, their computer self-efficacy, their attitudes toward risk, and their willingness to tinker. In this project the PI will exploit findings from gender research to develop and test, with the help of major industrial partners including IBM, Microsoft, Motorola and other industry members of the National Center for Women and Information Technology (NCWIT), a new methodology for software development that she hopes can make a real difference to the design of future systems so that females in technical problem-solving professions, such as computer science and other STEM fields, will gain access to software that no longer drains their capabilities but rather supports and advances them. If successful this research will, for the first time, enable software producers to find and remove subtle signposts in their systems that suggest "no girls allowed" with all the broad societal and workforce impacts that implies. To these ends, and especially focusing on problem-solving software, the PI will create a theory-based and empirically informed "gender lens" called GenderMaP (from GenderHCI Cognitive Walkthrough + Magnifying Persona) that will enable software designers and producers, even without a background in gender research, to identify ways in which their products may not be gender-inclusive. The resulting method will leverage theoretical and empirical gender difference foundations on motivation, computer self-efficacy, information processing styles, risk aversion, and willingness to tinker. The PI will investigate whether such a method can make a real difference to software's gender inclusiveness, by iteratively informing and empirically evaluating the method through at least 6 field studies with her industrial partners, and she will supplement the field studies with small-scale interviews, diary-like studies, and lab pilots. The investigation will be guided by, and contribute findings for, four research questions. RQ1 considers obstacles and remedies to adopting the method. RQ2 investigates how the gendered personas component of the method can be most effective in real use in industry, and how the personas interact with organizational needs and even stereotype bias. RQ3 investigates validity and other aspects of the overall value pay-off that the industrial partners gain from the method. Finally, RQ4 investigates the method's generality.