Effects of Visual Representation on Social Influence in Computer‐Mediated Communication

Two experiments investigated if and how visual representation of interactants affects depersonalization and conformity to group norms in anonymous computer‐mediated communication (CMC). In Experiment 1, a 2 (intergroup versus interpersonal) × 2 (same character versus different character) between‐subjects design experiment (N= 60), each participant made a decision about social dilemmas after seeing two other (ostensible) participants’ unanimous opinions and then exchanged supporting arguments. Consistent with the Social Identity model of Deindividuation Effects (SIDE), when the group level of self‐identity was rendered salient in an intergroup encounter, uniform virtual appearance of CMC partners triggered depersonalization and subsequent conformity behavior. By contrast, when the personal dimension of the self was salient, standardized representation tended to reduce conformity. To elucidate the mediation process, Experiment 2 investigated the causal links between depersonalization, group identification, and conformity. The results show that depersonalization accentuated adherence to group norms, both directly and indirectly via group identification.     

The Multiple Source Effect and Synthesized Speech

Two experiments examine the effect of multiple synthetic voices in an e‐commerce context. In Study 1, participants (N= 40) heard five positive reviews about a book from five different synthetic voices or from a single synthetic voice. Consistent with the multiple source effect, results showed that participants hearing multiple synthetic voices evaluated the reviewed books more positively, predicted more favorable public reaction to the books, and felt greater social presence of virtual speakers. The effects were mediated by participants’ feelings of social presence. The second experiment (N= 40) showed that the observed effects persisted even when participants were shown the purely artificial nature of synthesized speech. These results support the idea that characteristics of synthetic voices in doubly disembodied language settings influence participants’ imagination of virtual speakers, and that technological literacy does not hinder social responses to anthropomorphic technologies such as text‐to‐speech (TTS).     

How help during pregnancy can undermine self-efficacy and increase postpartum intentions to quit

Pregnancy reflects a common experience for women in today's workforce, yet recent data suggest that some women scale back or leave the workforce following childbirth. Considering these effects on women's careers, researchers have sought to understand the underlying dynamics of these decisions. Here, we explore a paradoxical reason for weakened postpartum career attitudes: help that women receive during pregnancy. We integrate stereotype threat and benevolent sexism theories to explain how the effects of help on postpartum intentions to quit may be transmitted through reductions in work self-efficacy. In doing so, we consider the role of perceived impact—or the extent to which help interferes with versus enables women's perceived ability to continue performing their work role. Results of a weekly diary study of 105 pregnant employees suggest that work-interfering help led to decreased self-efficacy for work during the following week. Furthermore, there was an indirect effect of average help received at work during pregnancy on postpartum intentions to quit the workforce through reductions in work self-efficacy that was stronger insofar as help was work-interfering versus work-enabling. Taken together, our results highlight unintended negative consequences that occur when others provide ineffective support to women at work during pregnancy.     

Perceptual Categorization: Connectionist Modelling and Decision Rules

Although it is currently popular to model human associative learning using connectionist networks, the mechanism by which their output activations are converted to probabilities of response has received relatively little attention. Several possible models of this decision process are considered here, including a simple ratio rule, a simple difference rule, their exponential versions, and a winner-take-all network. Two categorization experiments that attempt to dissociate these models are reported. Analogues of the experiments were presented to a single-layer, feed-forward, delta-rule network. Only the exponential ratio rule and the winner-take-all architecture, acting on the networks' output activations that corresponded to responses available on test, were capable of fully predicting the mean response results. In addition, unlike the exponential ratio rule, the winner-take-all model has the potential to predict latencies. Further studies will be required to determine whether latencies produced under more stringent conditions conform to the model's predictions.