اثر "جنسیت" کامپیوتر بر نفوذ اجتماعی اطلاعاتی: نقش تعدیل کنندگی در نوع کار

The present experiment examined if and how “gender” of the computer, manifested in character representation, would affect its informational influence on individuals’ decisions on masculine (sports) or feminine (fashion) topics. In a 2 (participant's gender) ×2 (character gender) ×2 (nature of topic: masculine vs. feminine) between-subjects experiment, participants played a trivia quiz game with the computer. During the game, they were given a chance to change their initial answer after seeing the computer's answer, which they knew was not necessarily correct. Results supported the match-up hypothesis such that while the male computer elicited greater conformity on the masculine topic than on the feminine topic, the opposite was true for the female counterpart. In addition, men were less likely to yield to the computer's suggestion than women on the masculine topic whereas women were less likely to succumb to the computer's influence on the feminine topic. These findings are discussed in terms of the robustness of gender-stereotyping in human–computer interaction and the implications for Computers Are Social Actors paradigm.

Unreasonable as it may sound, recent research on human–computer interaction has demonstrated that people reveal gender stereotypical reactions to the computer when its “gender” is manifested in the voice output—both male and female users rated a female-voiced computer as more knowledgeable about love and relationships than a male-voiced computer whereas a male-voiced computer was rated as more knowledgeable about technical subjects than its female counterpart (Nass et al., 1997). Even when the computer uses a clearly machine-generated synthetic voice, as opposed to recorded human voices, this gender stereotyping seems to persist: The product descriptions delivered by a male-voiced computer were considered as more credible than those by a female-voiced computer (Morishima et al., 2001); a male-voiced computer was perceived to be more competent and trustworthy and furthermore, elicited greater conformity to its recommendations than the female counterpart did (Lee et al., 2000). Taken together, the findings are in line with previously documented gender differences in credibility perception and social influence (see Carli, 2001; Ridgeway, 2001, for reviews), thereby supporting the notion that people respond to computers as they do to real people, albeit not consciously so (Nass and Moon, 2000; Reeves and Nass, 1996).