The present research investigated three approaches to the role of norms in the theory of planned behaviour (TPB). Two studies examined the proposed predictors of intentions to engage in household recycling (Studies 1 and 2) and reported recycling behaviour (Study 1). Study 1 tested the impact of descriptive and injunctive norms (personal and social) and the moderating role of self-monitoring on norm-intention relations. Study 2 examined the role of group norms and group identification and the moderating role of collective self on norm-intention relations. Both studies demonstrated support for the TPB and the inclusion of additional normative variables: attitudes; perceived behavioural control; descriptive; and personal injunctive norms (but not social injunctive norm) emerged as significant independent predictors of intentions. There was no evidence that the impact of norms on intentions varied as a function of the dispositional variables of self-monitoring (Study 1) or the collective self (Study 2). There was support, however, for the social identity approach to attitude-behaviour relations in that group norms predicted recycling intentions, particularly for individuals who identified strongly with the group. The results of these two studies highlight the critical role of social influence processes within the TPB and the attitude-behaviour context.

Most experiments on conformity have been conducted in relation to judgments of physical reality; surprisingly few papers have experimentally examined the influence of group norms on social issues with a moral component. In response to this, participants were told that they were either in a minority or in a majority relative to their university group in terms of their attitudes toward recognition of gay couples in law (Expt 1: N = 205) and a government apology to Aborigines (Expt 2: N = 110). In both experiments, it was found that participants who had a weak moral basis for their attitude conformed to the group norm on private behaviours. In contrast, those who had a strong moral basis for their attitude showed non-conformity on private behaviours and counter-conformity on public behaviours. Incidences of non-conformity and counter-conformity are discussed with reference to theory and research on normative influence.

PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN (in press)This research was supported by a fellowship from the Royal Netherlands Academy of Arts and Sciences to the first author, and a NWO Visiting Professorship grant to the second author. We thank two anonymous reviewers for their comments on a previous draft of the manuscript.Address correspondence to Jolanda Jetten at the School of Psychology, University of Exeter, Exeter EX4 4QG, United Kingdom. E-mail: j.jetten@exeter.ac.uk Peripheral Group Membership 2 Abstract Two studies investigated how both degree of identification and the individual's position within the group influence aspects of group loyalty. We considered ingroup position both in terms of the individual's current position within a group and expectations concerning the likelihood that one's position might change in the future. Peripheral group members learned that their acceptance by other group members would improve in the future or that they could expect rejection by other group members. Various indices of group loyalty (ingroup homogeneity, motivation to work for the group, and evaluation of a motivated group member) showed that when group members anticipated future rejection, the lower the identification, the less loyal they were. In contrast, those who expected future acceptance were more loyal (more motivated to work for the group) the lower their identification. Current group behavior depends on both intragroup future expectations and level of identification. Peripheral Group Membership 3Predicting the Paths of Peripherals:The Interaction of Identification and Future Possibilities Social psychologists have made important advances in understanding group behavior, particularly that of "prototypical" group members (e.g., Jetten, Spears, & Manstead, 1997;Moreland, 1985;Moreland, Levine, & Cini, 1993). Prototypical group members are more likely to be group leaders (Eagly, Makhijani, & Klonski, 1992;Hains, Hogg, & Duck, 1997), successful in eliciting attitude change in others (van Knippenberg, Lossie, & Wilke, 1994), evaluated more positively than other group members (Hogg & Hardie, 1991), and they are more likely to define the group's norms and act in accordance with those norms (Oakes, Haslam, & Turner, 1999;Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). The present research, in contrast, focuses on group members who are not, or are not yet, perceived as prototypical of the group.Though less often investigated, peripheral or "non-prototypical" group members are potentially more intriguing because their behaviors have proven to be more difficult to predict compared to that of prototypical members (Schmitt & Branscombe, 2002).Peripherals who are less typical or central to the group, are often more variable in their responses than prototypicals. Peripheral status can lead people to deviate from group norms (e.g., Lewin, 1948), but it can also encourage people to attempt to satisfy the norms of the group even more strenuously (e.g., Breakwell, 1979; Noel, Wann, & Branscombe,1995;Tajfel, 1978). So...

Based on the assertion that previous research may have inadvertently confounded two stereotypes, we considered the impact of benevolent sexism on rape victim blame in the context of independent manipulations of gender and the perceived genuineness (victim stereotypicality) of an acquaintance rape victim. We predicted that for blame, benevolent sexism may be independently positively associated with gender and victim counter-stereotypicality. Following pilot work, 120 Australian undergraduates read an acquaintance rape scenario. Results indicated that benevolent sexism was only positively associated with blame of the gender counter-stereotypical victim when that victim was also counter-stereotypical in terms of victim stereotypes. This result indicates a more moderate role than previously indicated for benevolent sexism in accounting for rape victim blame.

Jurors rely on a range of schemas when evaluating allegations of rape and sexual assault. For example, they may be influenced by the prototypicality of the alleged offense, the stereotypicality of the victim, or gender-related stereotypes. These schemas have often been conflated however, making it difficult to determine the unique impact of each on jurors' perceptions. To be able to effectively counter any schema-related misconceptions, we must first identify which beliefs are important and when. An experiment (N = 420) examined the independent effects of offense prototypicality and victim stereotypicality on mock jurors' perceptions. As expected, victim stereotypicality had a greater effect on judgments in the counter-prototypical (acquaintance) assault scenario than in the prototypical (stranger) assault scenario. When the complainant was described as being a counter-stereotypical victim in the acquaintance rape scenario, the defendant was seen as less likely to be guilty and evaluated more positively and the complainant less positively compared with when the complainant was described as being a stereotypical victim. Analysis of the qualitative data suggested a focus on different factors in reaching verdicts in the stranger and acquaintance rape scenarios. Results were interpreted as evidence that jurors "step down" through a hierarchy of schemas in their attempts to determine what happened in cases of rape and sexual assault.

The impact of social support on dissonance arousal was investigated from a social identity view of dissonance theory. This perspective is seen as augmenting current conceptualizations of dissonance theory by predicting when normative information will impact on dissonance arousal and by indicating the availability of identity-related strategies of dissonance reduction. An experiment was conducted to induce feelings of hypocrisy under conditions of behavioral support or nonsupport. Group salience was either high or low, or individual identity was emphasized. As predicted, participants with no support from the salient in-group exhibited the greatest need to reduce dissonance through attitude change and reduced levels of group identification. Results were interpreted in terms of self being central to the arousal and reduction of dissonance.

Two experiments provided support for the central hypothesis--derived from social identity/self-categorization theories--that attitudes would be most likely to predict behaviour when they were supported by a congruent in-group norm. In the first experiment, norm congruency and mode of behavioural decision-making (spontaneous or deliberative) were orthogonally manipulated in a between-subjects study of career choice in psychology. Participants exposed to an attitudinally congruent in-group norm towards their preferred career choice were more likely to display attitude-behaviour consistency than those exposed to an attitudinally inconsistent group norm, an effect that was evident under both spontaneous and deliberative decision-making conditions. Using a mock jury paradigm, Expt 2 replicated and extended the first experiment by including a manipulation of in-group salience. As predicted, participants exposed to an incongruent norm displayed greater attitude-behaviour inconsistency than those exposed to a congruent norm. Contrary to predictions, this effect did not vary as a function of group salience, nor did the effects of group norms for high and low salience participants vary as a function of mode of behavioural decision-making. However, there was evidence that perceived identification with the group moderated the influence of norms on attitude-behaviour consistency.

Te sexual assault victim "who comes to the attention of the authorities has her victimization measured against the current rape mythologies" (R v. Seaboyer, 1991). Tis is particularly troubling given that lay beliefs regarding the crime of sexual assault are at odds with the data documenting the circumstances surrounding actual rape. Research has consistently demonstrated that lay people (hence, jurors) will question the validity of a sexual assault claim and judge the victim more harshly, if the circumstances surrounding the assault and/or the characteristics and actions of the sexual assault complainant do not comport with people's expectations about the event. In this paper we report the results of a juror simulation that examines the impact of victim's postassault emotional demeanor on judgments, in the context of independent manipulations of gender stereotypicality and victim stereotypicality. Results revealed that the complainant's emotional display had a powerful impact on participants' judgments, with the claim viewed as more valid when the complainant was portrayed as tearful/upset as opposed to calm/controlled, but only when the complainant was portrayed as gender stereotypic.