Affect toward Computers Who Coerce in Social Exchange.

How does interaction with computer or human actors change one's affective perception of them? Specifically, how do coercive interactions with computers or humans alter such impressions? "Computers Are Social Actors" (CASA) is human-computer interaction research from a psychological and communications tradition that posits that people interact with computers in fundamentally social ways. CASA research has been primarily concerned with how people interact with one computer differently from another, not how the interactions differ between computers and humans. Social exchange theory suggests how coercive behaviors alter affective perceptions of exchange partners, and affect control theory (ACT) predicts how perceptions of actors who coerce differ based on their identity as a computer or human and based on the observer's gender. I conduct an experiment modeled on a previous study by Linda Molm in coercive, reciprocal social exchange. I modify and extend this study by examining how exchange partner's strategy (coercive or non-coercive), partner's identity (human or computer), and subject's gender influence the subject's affective perception of her partner. The results support CASA literature with computer and human identities not differing in perceived affect, except for coercive computers who have greater perceived goodness compared to coercive humans. All coercive actors are perceived as less good by females than by males. [ABSTRACT FROM AUTHOR]