This study examined the relationship between consequence probabilities and decisions concerning self-protective behavior. The hypothesis put forth was that the probabilities associated with many common risks are at a level at which variations in probability have little influence on risk decisions. Subjects were presented with hypothetical risk situations in which consequence probability, consequence severity, and the costs of precautionary actions were explicitly stated and systematically manipulated. Decisions concerning the likelihood of taking a specific precautionary action in response to these risks were examined as a function of these three factors. The results revealed a "probability equivalence" effect, ranges of probabilities over which precautionary intent was essentially invariant, and that these ranges were highly dependent on consequence severity. In general, precautionary intent decreased as a function of decreasing consequence probability. However, for situations with low consequence severity, precautionary intent began to stabilize, trending towards an apparent minimum at low probability levels. Increasing consequence severity also had the effect of decreasing the "probability threshold", the probability level below which subjects were unlikely to take precautionary action. Increasing precautionary costs was also found to consistently depress willingness to take precautions. Precautionary intent was also examined as a function of several characteristics of the decision makers. Most scenarios revealed that levels of precautionary intent increased with age. In addition, in more than half of the scenarios, males were less inclined to take precautions than females. Analyses at the individual subject level revealed considerable variability in probability threshold and equivalence effects, indicating that the use of probabilities in risk decisions is highly dependent on the characteristics of the decision maker and the risk situation. Collectively the findings support the notion that probability information can influence decisions concerning precautionary intent, but that the probability equivalence effect may render consequence probability a poor discriminator when judgments are made concerning similar risks. The results suggest that the probability effects were products of subjects' informal cost-benefit analyses as opposed to a failure to perceive meaningful differences in the probabilities.

Evidence cited as validating the Vroom-Yetton model has relied on managers' self-reports of the behavior they did or would exhibit. Research suggests that knowledge of group performance can bias both group process and leadership ratings. An experimental study was conducted to determine whether descriptive ratings of leader decision-making behavior are influenced by performance cues. Further, self-ratings of their behavior biased toward consistency with implicit theories of performance. Leader decision-making style & level of performance feedback were manipulated in a 2 x 3 factorial design. Leaders' self-ratings were influenced by high performance feedback toward consistency with their implicit theories of performance. When given high performance feedback, group members rated their leader as more participative than a non-interacting observer, but only in the participative leader condition. Contrary to prior research, group process ratings were not biased by performance feedback. Evidence suggesting that rating leader behavior influences ratings of group process is presented.

Little is known about how groups use their most precious commodity: information held by group members. In particular, until the recent work of Stasser and his colleagues, almost nothing was known about how groups differentially use commonly held information and information that is known only by one member. Stasser's probabilistic information sampling model explains differential treatment of uniquely and commonly held information by explaining that more group members have access to commonly held than uniquely held information, so the commonly held information is more likely to be mentioned in discussion (Stasser & Titus, 1985, 1987). This model has been primarily supported for first mentions of information at the group level (Stasser, Taylor, & Hanna, 1989). In the study reported here predictions from a suggested model were tested on second and further mentions of information to determine if anything other than probability contributes to withholding of unique information by group members. It was expected that group member concerns about utility of information (including task relevance and validity) and social norms regarding sharing unique information might contribute separately and in combination to the withholding of unique information after it was discovered unique (via the first mention). Manipulations of confirmation of information utility and social norming were expected to increase repeat mentions of uniquely held information relative to commonly held information. University students read information about study abroad programs and decided as a group whether to recommend such a program for their university. Conversations of two- and three-person groups were audio-tape recorded and analyzed at both the group and individual level for mentions of commonly and uniquely held information. Perceptions of information usefulness, recognition of item uniqueness, and perceptions of group process were gathered after the group discussion. As expected, the manipulations did not affect first mentions. For second and further mentions social norming and utility confirmation singly and in combination tended to lower the advantage of commonly held information, although not always significantly. Suggestions for further research are made and recommendations for applications for decision making groups are based upon the demonstrated positive effectiveness of the utility confirmation manipulation.

This study investigates the effects of perceived morality and facial attractiveness on jurors' allocations of responsibility in a civil litigation context. Confronted with uncertainty, jurors may focus less on the factual evidence presented and more on peripheral cues. It was hypothesized that jurors will attribute more responsibility to a bad individual compared to a good person, and unattractive consumers will be allocated more responsibility for an accident than their attractive counterparts. Results indicate perceived morality significantly influences responsibility allocations. Bad people were given more responsibility than good people. Although unattractive consumers were given more responsibility than attractive consumers the difference was not statistically significant. Finally, allocations to the consumer and manufacturer differed depending upon the products used in each scenario, indicating that responsibility allocations to consumers are related to the type of product. Implications for civil litigation include the awareness that jurors decisions are influenced by the type of product and subjective information about the consumer.

The terms "public" and "private" are useful in describing the extremes ot human experience. Unfortunately, actual human experience is more complex than a simple dichotomy can convey. Are we public or private when making a telephone call from a phone booth. Neither? Both? A more apt model is that of a matrix of public and private experiences and spaces. At home watching television, we may be in a private space, but the experience has public overtones in that it is simultaneously shared by others. In a religious ceremony, on the other hand, we have private experiences, even when surrounded by others in a public space. Private experiences in public spaces are awkward for most of us. When the experience is mediated, however, it can take on positive qualities. Standing on line in a cafeteria (alone in a crowd) should be uncomfortable, but that feeling is mitigated by movement along a sequence of thresholds, gradations of space, and events. Despite initial discomfort, we become more comfortable in a communal experience. This project applies these ideas to a program at the scale of a community, in this case, a retirement community. For most people, the move into a retirement community is dreaded as representing a loss of freedom and individuality. This thesis's strategies of mediation attempt to mitigate the anxieties of a new environment, while still offering a varied and vital place for living.

Recent research suggests that displays of discrimination have changed from expressions that were once very overt in nature to expressions that are more subtle and involve interpersonal behaviors (Hebl, Foster, Mannix, Dovidio, 2002). Unlike formal discrimination, there is little to no legislation that protects stigmatized individuals from interpersonal discrimination; hence, research on strategies for remediating interpersonal discrimination is becoming increasingly important. The current research examines three strategies (compensation, acknowledgment, and individuating information) that attempt to reduce the interpersonal discrimination that gay men and lesbians receive in a job applicant setting. Results show that employing a remediation strategy involving compensation or acknowledgment reduces the interpersonal discrimination that stigmatized individuals receive when applying for jobs.

Previous research on mediation has not examined the influence that disputant affect state may have on perceptions of mediators and proposals for resolving the conflict. This influence may be particularly important in situations where a mediator has similar interests, or is aligned, with one particular disputing party. The present study attempted to examine how affect states influence the evaluations of mediators with unfavorable and favorable alignment. Hypotheses were that, compared to neutrals, negative and positive affect states would lead to increased reliance on heuristics such as mediator alignment for determining perceptions of the mediation situation. One hundred twenty-one participants wrote essays designed to induce angry, neutral, or happy affect states, and then participated in a mediation task. Results indicated no support for the hypotheses. The lack of results may be attributed to insensitivity of the affect and alignment manipulations. Implications for future research are discussed.

This research proposes a revised conceptualization of the construct of diversity in groups to include both the traditional, objective diversity as well as perceptual, subjective diversity. Theoretical rationale for the differentiation is discussed and a framework is proposed in which objective and subjective diversity both have independent effects on group processes and outcomes. Hypotheses were tested through the use of a laboratory experiment in which 46 groups of 3--4 persons each performed organizational simulation tasks. Results include a series of hierarchical regression analyses that provide initial support for the differentiation between objective and subjective diversity. Issues of aggregation, level of analysis, and methodology are discussed. Additionally, the implications of the differentiation of objective and subjective diversity are highlighted.

The current study focuses on understanding how individuals perceive bias. Using fictitious candidates for a job in differently discriminating companies, this study investigated three sets of factors hypothesized to contribute to the perception of bias; characteristics of the target job candidate, characteristic of the context, and characteristics (attitudes) of the participant. Participants were individuals approached in a Houston airport waiting lounge who returned the survey via a postage paid envelope (n = 108) or Rice University undergraduates (n = 49) participating in exchange for course credit. Investigation of the target characteristics showed that the target's deservedness (the extent to which the target's qualifications matched the target's selection outcome) greatly determined perceptions of bias. Race of the target was found to be more complexly related to perception of bias than originally hypothesized. The notion of target deservedness was also used to investigate the context and participant factors as contributions to bias. The theory of shifting standards was used to predict that context (past company discrimination in the current study) would influence bias in a counter-intuitive direction such that bias would not be perceived where it was expected. Perceptions of bias supported the hypothesis for deserved targets, but the results were counter to the hypothesis for undeserved targets (with the exception of selected white candidates). Participant attitudes were investigated as a third factor. Participant attitudes were not found to be predictive of bias perceptions. Alternative analyses suggest that participant attitudes may be useful for explaining bias perception when outcomes are undeserved.

The purpose of the present study was to examine the extent to which men and women in a group internalize group feedback, and what mechanisms might underlie this internalization. Task/gender congruence, perceived contribution, and collectivism were all proposed contributors to gender differences in feedback internalization, and the consequent relationship between self- and collective-efficacy beliefs. Participants completed a brainstorming class and were given false group feedback. Results indicated that there were gender differences in response to the task-oriented condition. Also, perceived contribution moderated the relationship between feedback and self-efficacy, and between self-efficacy and collective-efficacy.