Date of Award

8-2018

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Curriculum & Instruction

Committee Chair

William Watson

Committee Member 1

Marisa Exter

Committee Member 2

Anne Traynor

Committee Member 3

David Whittinghill

Abstract

The underrepresentation of women in computer science is a serious issue with ramifications that affect not only women working in the field, but also the field at large and the national economy. While the problem is attributed to several interconnected causes, such as computer knowledge, experience, and familiarity; they do not tell the entire story, nor do they point to a global trend. Studies indicate that these factors did not stop the computer science field from achieving a balanced male-female ratio in non-western countries. Since no single factor can be attributed as the primary cause of this complicated problem, and in an effort to minimize the issue and move computer science in western countries closer toward an egalitarian model, the approach taken to enhance the representation of women in computer science needed to be multifaceted. The most common factors revealed in the literature as possible contributors to the problem were listed and categorized. Based on these factors, the study hypothesized a model (Amal’s Model) and tested its effectiveness at predicting students’ domain identifications with computer science. Following this, the study suggested and tested the impact of coupling constructionist gaming with studio pedagogy in a game design studio on students’, especially women’s, learning of, self-efficacy in, attitudes toward, and domain identifications with computer science. The results compared the implementation of the game design studio with robotics and traditional pedagogical practices.

The participant body (N=94) was composed primarily of computer science majors and non-majors who were enrolled during the Fall 2017 semester. The research design included block randomization in order to make sure that the male-female ratio was relatively balanced across all of the groups. A pre-posttest experimental design was utilized to compare students’ learning of, self-efficacy in, attitudes toward, and domain identifications with computer science in three groups: game design studio, robotics and coding, and a control group, with special attention to any gender-based differences that were revealed during the study. The data were analyzed using statistical tests and results showed that Amal’s Model was significantly effective at predicting students’ (both men’s and women’s) domain identifications with computer science. The study showed that aspects of Amal’s Model, including students’ learning of, self-efficacy in, and attitudes toward computer science, were significant predictors of their computer domain identifications with the field. The results concluded that increasing women’s domain identifications with the field, to match domain identification levels held by men, would, in turn, encourage women to pursue a degree in the field.

Results from the current study showed that the game design studio had a significant impact on students’ (both men’s and women’s) learning of, self-efficacy in, attitudes toward and domain identifications with computer science when compared with the control group. When robotics was added to the comparison, the game design studio was effective at improving students’ learning and attitudes, and significantly effective at improving their self-efficacy and domain identifications from the pre-test to the pos-test. Unlike the findings produced by several other studies conducted in this area, when students in the study were asked about the suitability of the field for women, the majority of women and men involved agreed that the field is suitable for women.

In analyzing the results for apparent gender-related changes in students’ learning, self-efficacy, attitudes, and domain identifications from pre-test to the post-test, it was found that both teaching approaches were effective at improving men’s and women’s learning from the pretest to the post-test. Regardless of the differences between groups in men’s learning of, self-efficacy in, attitudes toward, and domain identification with the field, the differences were statistically insignificant across the groups. While women in the robotics and coding group had significantly higher self-efficacy beliefs than the control group, women in the game design studio showed significant improvements, not only in their self-efficacy, but in their learning, attitudes and domain identifications when compared with the control group. Even though women in the game design studio had slightly lower domain identification levels than men in the same group, they demonstrated the highest statistical improvement in their domain identification with computer science when compared to not only the control group, but also the robotics and coding group. The study demonstrated that women in the control group consistently had the lowest scores in their learning of, self-efficacy in, attitudes toward, and domain identifications with computer science. The results, including current findings and recommendations for enhancing the representation of women in computer science, and suggestions for future studies, were discussed in great detail.

Download

Share

COinS