FIE2013

gistfile1.txt

We report on the experience of teaching an elective, pilot course on Artificial Intelligence (AI) in Computer Games within the Simulation and Game Design department at a two-year community college during a 16-week Fall 2012 semester. An advisory committee comprising both academic and industry members added the course to address two primary concerns: 1. to evaluate if teaching through game design can positively increase attitudes about Computer Science, and 2. to assess alternative pedagogical approaches, which include a blended classroom model, active learning techniques, and the use of practical, computer-based coding exercises and exams in lieu of traditional multiple-choice or true-false formats. We selected Python as the course programming language, based on suggestions from industry partners and academic literature. The course content can be summarized as two weeks of learning the Python language features, an additional two weeks reviewing mathematical foundations, and the remaining semester time on AI game topics. To identify changes in student attitudes, students completed the Moskal's Attitudes Toward Computer Science survey (n = 10) twice during the semester, once immediately before the core AI component of the course and again at the end of the semester. The results indicate that student attitudes positively and significantly increased during the semester in both professional (p = 0.037) and interest constructs (p = 0.014). To evaluate instructional compatibility, students completed a Felder-Soloman Index of Learning Styles (ILS) questionnaire (n = 14) and we compared these results against a student population from an Introduction to Computer Science course at a four-year University. We were unable to identify any statistical differences between the two populations, which suggests that successful techniques based on ILS from four-year institutions can potentially be applied at the community college level. Students also completed an an end-of-class evaluation (n = 13) with 5-point Likert-item questions to gauge student perceptions. 10 students (77%) considered the practical coding exam to be appropriate for the course. Moreover, 10 students (77%) preferred to use the same exam format for future courses. 10 students (77%) indicated that the course was either extremely important or very important in obtaining an entry-level game developer position. 8 students (61%) very strongly preferred or strongly preferred the blended classroom model of the course. 12 students (92%) reported that the allocated 5-10 minute one-on-one progress checks with the instructor were very useful or useful. 8 students (62%) reported that they would be very likely or likely to use Python as their preferred language in future programming tasks. We are encouraged by these results. Finally, we performed semi-structured interviews with local game companies. The interview results show that companies are favorable to the course and that the material is appropriate for the entry-level industry scripting and game-testing positions typically pursued by our graduates. The results of this work serve as a template for community colleges considering the adoption of specific pedagogical approaches in the classroom or for colleges who wish to adopt a prepared AI course in its entirety.