Arroyo, I., Woolf, B.P., Royer, J.M., Tai, M. (2009) Affective Gendered Learning Companion, International Conference on Artificial Intelligence and Education, Brighton, England, IOS Press 1 Affective Gendered Learning Companions Ivon ARROYO, a Beverly Park WOOLF, a James M. ROYER, b Minghui TAI c a Department of Computer Science, b Department of Psychology c School of Education University of Massachusetts Amherst Abstract. We researched the impact of gendered pedagogical agents on student attitudes for math, motivation and achievement in math, within the context of an adaptive tutoring software for high school mathematics. Learning companions emphasize perseverance by valuing effort in challenging tasks. They are also empathetic, as they reflect students’ emotional states. The results suggest that, across two studies, it was the male learning companion that produced the most positive impact on female students’ state-based emotions, attitudes and learning. It is possible that girls transfer their stereotypes to the computer software. Keywords. Intelligent tutors, gender differences, pedagogical agents, emotions, motivation and affect, evaluation, mathematics education Introduction Student learning is one of the most important outcomes of education, and research has increasingly documented the importance of emotions and attitudes in learning. In the case of mathematics, research (summarized in [1]) has shown that the subjective value of math and students’ self-concept of math ability drops as students transition from elementary school to junior high school. In addition, there is a sharp rise in the perceived difficulty of math around 7 th grade and persisting through grade 12, during which time subjective value of math and math self-concept declines steadily, and during which time, perceived difficulty of math steadily increases. These results are in contrast to the pattern for English learning which do not show grade related changes. The shift in attitude about the value of math skills and student self-concept of math ability is particularly apparent among females and minority groups [2, 3]. Changes in math self-esteem among females are disassociated from indicators of math performance. In a very large study involving thousands of students, Catsambis [2] showed that 8 th grade females in three ethnic groups were similar to their male counterparts in math test performance, generally better in math grades, and were taking more higher level math courses. Despite these objective indicators of performance, females from all three ethnic groups reported less interest in math intensive careers than their male peers, and showed less inclination to look forward to math class. These attitudinal gender differences are undoubtedly related to the fact that females transitioning into higher education chose math intensive career paths with less frequency that their male counterparts. This is a problem of concern in the United States where, for instance, women earn only 21% of both BS and MS degrees in engineering, and 25% and 31% of computer science BS and MS degrees, with lifelong implications for women’s earnings [4].