The social identity approach to attitude-behaviour relations (Terry & Hogg, 1996) argues that the relationship between attitudes and behaviour will be strengthened if individuals perceive normative support from an important reference group, an argument supported by both field and experimental research. Most notably, the approach not only provides a valuable perspective on the question of when attitudes and norms will influence behaviour, but also on the question of for whom norms will be most influential. However, recent research has suggested that it is not only individuals high in identification who display group behaviour (e.g., Barreto & Ellemers, 2000a, 2002). Within the social identity approach, there is increasing recognition that individuals might display group behaviour for strategic reasons, such as the desire for positive evaluations or to improve one's position in the group, as well as for more group oriented reasons. In
light of the likely impact of strategic motivations, it is important to examine such effects in the context of attitude-behaviour relations. The six studies reported here build upon and extend the social identity approach by examining whether strategic motivations impact upon group-mediated attitude-behaviour consistency.
Studies 1 to 3 examined the impact of anonymity and accountability on the attitude-behaviour relationship. In all three studies, participants completed attitude measures, reported their identification with the group or salience of group membership was manipulated, were exposed to ingroup normative information, and completed dependent measures under the expectation of complete anonymity or accountability. Study 1 (N= 93) investigated the interplay of identification and accountability under conditions of high ingroup normative support. High identifiers displayed greater attitude-behaviour and attitude consistency under
anonymous conditions, but low identifiers reported greater consistency when accountable to the ingroup. In Study 2 (N = 213), salience of social identity, ingroup normative support, and accountability were manipulated. As in Study 1, low salience participants displayed more group normative behaviour in accountable response conditions. High salience participants either displayed more group normative behaviour when anonymous (on the willingness measure), or responses did not vary across response context. In Study 3 (N = 124), the impact of ingroup versus outgroup accountability was compared. Low identifiers displayed group-mediated behaviour only when accountable to an ingroup audience, suggesting that low identifiers adapt to audience expectations rather than group expectations. High identifiers displayed more attitude-consistent behaviour to an outgroup audience.
Studies 4 and 5 investigated the role of position within
the group. In both studies, participants reported their attitudes and identification, were exposed to ingroup normative information, received false feedback about their intragroup position, and completed the dependent measures. In Study 4 (N= 130), in which all responses were anonymous, only high identifying core group members exhibited group normative behaviour. High identifying peripheral members, and low identifiers, did not display group behaviour. Study 5 (N= 126), examined the strategic behaviour of peripheral members further by including a manipulation of accountability. In anonymous contexts, the responses of low and high identifying peripheral members did not differ. However, when accountable, low and high identifying peripheral members differed in their displays of group behaviour with high identifying peripheral members displaying more group normative behaviour than low identifiers.
Study 6 examined social
identity and strategic processes outside the experimental setting. Respondents (N = 187) completed measures of attitude, subjective norm, perceived behavioural control, ingroup norm, and ingroup identification. Scenarios describing consumption of one's preferred beer in the presence of either an ingroup or an outgroup audience were used to manipulate type of audience. The dependent measures were intentions to consume, likelihood of consumption in the scenarios, and perceived audience reactions to consumption. As expected, the effect of norms was moderated by identification and type of accountability—level of normative support influenced likely consumption and perceived reactions only for the high identifiers in response to the ingroup audience scenario. Moreover, the effect of normative support on the behaviour of high identifiers was mediated through the perceived audience reactions, suggesting the operation of strategic processes.
Overall, the six studies provided some evidence that displays of group-mediated attitude-behaviour consistency are the product of a strategic consideration of the level of ingroup normative support, attachment to the group, position within the group, and the degree of accountability.