Deity Representation: A Prototype Approach

This research systematically evaluates via prototype analysis how conceptualizations of Western adult's monotheistic God are structured. Over 4 studies, using U.S. student and community samples of predominantly Christians, features of God are identified, feature centrality is documented, and centrality influence on cognition is evaluated. Studies 1 and 2 produced considerable overlap in feature frequency and centrality ratings across the samples, with “God is love” being the most frequently listed central feature. In Studies 3 (choice latency) and 4 (recall and recognition memory), the centrality of features influenced cognitive processes: central features were more quickly identified as features of God than peripheral features; were correctly recognized more often; and central features were correctly recalled more often than peripheral features. Results indicated that participants meaningfully judged centrality and that centrality affected cognition. Thus, the two criteria necessary for demonstrating deity representations adhere to a prototype structure were met. Implications and future directions are discussed.

The Content of Self- and Other-Concepts: A Network Perspective

The predictive value of the network perspective on the content of person-concepts was investigated in two studies. A number of predictions were tested based on the positive relationship between person-concept feature centrality and accessibility derived from the network model. Consistent with a network account of person-concept feature centrality, it was expected that features spontaneously generated by subjects when asked to describe a particular person would be shown to be more accessible on independent indices of accessibility within the context of the same person relative to features that were freely generated for a different person target and nomothetic features. Study 1 attempted to demonstrate that yes-no person-referent descriptiveness judgments involving spontaneously generated features would be made more quickly relative to the other two categories of features. Utilizing an experience-sampling methodology, Study 2 attempted to show that spontaneously offered features would be applied more often to the person for which they were generated when thinking about him or her in ongoing thought. The findings in both studies were consistent with predictions. Because reaction-time and frequency measures are independent indices of accessibility, the convergent findings across the two studies provide compelling support for the predictive utility of the network model.