University of Groningen. Identity patterns in diverse workgroups Vos, Menno Wouter

Transcription

1 University of Groningen Identity patterns in diverse workgroups Vos, Menno Wouter IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2009 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Vos, M. W. (2009). Identity patterns in diverse workgroups: improving social integration outcomes through relational identities s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 Identity patterns in diverse work groups Improving social integration outcomes through relational identities

3 Cover: Printing: Megan Hoffman / Ridderprint B.V. Ridderkerk

4 RIJKSUNIVERSITEIT GRONINGEN Identity patterns in diverse workgroups Improving social integration outcomes through relational identities Proefschrift ter verkrijging van het doctoraat in de Gedrags- en Maatschappijwetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op donderdag 16 april 2009 om uur door Menno Wouter Vos geboren op 26 december 1978 te Amsterdam

5 Promotores: Beoordelingscommissie: Prof. dr. K.I. van Oudenhoven - van der Zee Prof. dr. A.P. Buunk Prof. dr. P. B. Paulus Prof. dr. T. T. Postmes Prof. dr. G. S. van der Vegt ISBN:

6 Contents Chapter 1 General introduction 7 Chapter 2 Chapter 3 Chapter 4 Individual differences in self representation: The development of a measure of personal, relational and collective identity orientation 19 Study Study With a little help from my friends: Enhancing prosocial behavior in diverse work groups by focusing on relational identities 41 Study Study Study Patterns of social identification in diverse work groups: The influence of collective and relational identities on team satisfaction 63 Chapter 5 Summary and discussion 79 References 89 Samenvatting (Summary in Dutch) 105 Dankwoord (Acknowledgments) 111

7

8 Chapter 1 General introduction During the preliminary elections in 2008 for the candidacy for the next president of the United States, the Democratic Party put forward two candidates who both had the potency and were eager to represent the democrats as a candidate running for presidency: Hillary Clinton and Barack Obama. Although both candidates shared democratic views, they differed in opinion on certain key matters, such as the military presence of US troops in Iraq and the health care system. This difference of opinion and vision on these key matters created a division within the Democratic Party, but also among democratic voters. Obama held the majority of the black votes and also appealed to the younger voters. Clinton on the other hand, had close ties to the Hispanic community and could count on strong support of the working class and (elder) women. Both candidates put much effort and resources in their political campaign to reach the majority of delegates needed to win the pre-elections. Eventually, Obama had gained enough delegates to become the presumptive nominee. In a speech before her supporters on June 7th 2008, Clinton ended her campaign and endorsed Obama. Although they had been rivals during the preliminary elections and had different views on political matters, both were strongly inclined to re-establish their relationship to reunite the Democratic Party. According to the Obama s chief strategist: They both had an interest in working closely together despite their different opinion on certain matters (Nagourney & Zeleny, 2008). The previous example shows that individuals from a shared collective in this case the Democratic Party - with different views and opinions may come into conflict with each other, resulting in a division within the collective. In this case, a successful reunion seemed to depend on the candidates strong orientation to (re-)establish positive relationships with each other. The present dissertation will focus on such a relational orientation as a strategy to reduce problems in groups which are characterized by dissension of views, values and identities. More specifically, we will argue that an identity which encompasses a strong focus on interpersonal bonds is beneficial in work teams in which people from different backgrounds (e.g. ethnicity, gender, and age) have to work together. Diversity in organizations Many organizations are now recognizing that a diverse workforce can be a valuable resource to gain competitive advantage. Also in academia the interest in diversity is increasing. Over

9 8 Chapter 1 the last decades many researchers from different domains have examined effects of diversity on group processes and performance (for an overview see Horwitz & Horwitz, 2007; Williams & O Reilly, 1998). These scholars have conducted laboratory and field studies examining performance and social integration outcomes such as identification processes and well-being as a function of group composition in terms of ethnicity, gender, age, personality or functional background. The literature on diversity predominantly distinguishes between social category diversity i.e. differences between individuals based on readily detectable (social) demographics, such as ethnicity, gender and age - and task related/informational diversity - i.e. diversity based on acquired, less visible individual attributes, such as functional expertise or education (Jehn, Northcraft & Neale, 1999; Milliken & Martins, 1996; Tsui, Egan & O Reilly, 1992; Van Knippenberg, de Dreu & Homan, 2004). Although it sometimes may be difficult to separate this distinction strictly in a diverse work context, as employees may differ from each other on a combination of characteristics on both dimensions, it seems warranted to make this distinction, as previous research has shown that outcomes differ dependent on the type of diversity dimension (e.g. Van Knippenberg et al., 2004). Many authors have claimed that diversity can be beneficial for organizations. Cox and Blake (1991) pointed out several competitive advantages of diversity among employees in organizations. The cognitive diversity hypothesis predicts that team diversity has a positive impact on performance because of unique cognitive attributes that members bring to the team (Cox & Blake, 1991; Hambrick, Cho, & Chen, 1996). Ultimately, cognitive diversity among heterogeneous members may promote creativity, innovation, and problem solving, and eventually results in superior performance relative to cognitively homogeneous teams. Studies from this perspective have indeed shown that diversity based on functional expertise or education promotes team performance (for an overview, see Horwitz & Horwitz, 2007). However, with regard to differences between individuals on social category dimensions, positive outcomes of diversity do not seem to be so evident (Horwitz & Horwitz, 2007). On the contrary, workplace diversity may sometimes to be difficult to manage as it goes along with different values, views, opinions and identities among employees which may not be compatible with each other. Subgroups may easily emerge with team members from the same background or with the same values, views and opinions sticking together, resulting in tensions within the team (Van Knippenberg et al. 2004). Potential benefits of diversity in terms of improved innovation, creativity and performance may be reduced due to internal conflict, dissatisfaction and lack of identification (Jehn, Chatwick, & Thatcher, 1997; Jehn et al., 1999). To preserve potential benefits of working in a

10 General introduction 9 diverse work context, it seems important to tackle these problems of social integration, especially since a socially well integrated team seems to be an important prerequisite to accomplish tasks effectively (e.g. Horwitz & Horwitz, 2007; Tuckman, 1965). Diversity and identities The general understanding of the underlying processes that lead to positive and negative diversity outcomes remains limited (Brickson, 2000). Although many approaches to understand diversity in an organizational setting have emerged, theorists generally agree that identification processes play a central role in this regard (Nkomo & Cox, 1996). Some authors have even incorporated identities into the very definition of diversity: a mixture of people with different (group) identities within the same social system (Nkomo & Cox, 1996; p. 339). Insights in the specific identities that are existent in diverse workgroups and how they relate to social integration outcomes may broaden theoretical knowledge about underlying processes that predict outcomes of team diversity but may also provide organizations means to manage diversity effectively. The most influential theory regarding identities in a social context is Social Identity Theory (SIT; Tajfel & Turner, 1979). SIT assumes that people enter social situations with a self-concept that is basically determined by their personal identity and their social identity (see Turner, 1982). Personal identity is regarded as the individuated self those characteristics that differentiate one individual from others within a given social context. Personal identities are contrasted with social identities, which reflect categorizations of the self into more inclusive social units that depersonalize the self-concept (Tajfel & Turner, 1986). When social identities are salient, individuals have a tendency to categorize themselves as members of social groups, and the characteristics that belong to the group will be internalized in the self. More recently, Brewer & Gardner (1996) expanded SIT s notions on personal and collective identities by introducing an extended model of the self in which a self view based on an individual s conception of their relatedness to other individuals, was added. When a relational identity orientation is salient, individuals conceive themselves in terms of relationships with other individuals. All three identity orientations seem to be relevant in an organizational context (see Brickson, 2000; Brickson & Brewer, 2001). As a personal identity orientation reflects a strong focus on one s own unique characteristics and attributes, this identity orientation may promote interpersonal competition, in which an individual s performance is compared to the performance of his/her fellow co-workers. A collective identity orientation

11 10 Chapter 1 will likely trigger a sense of belongingness to a social category, for example the work team, department or organization as a whole. A relational identity orientation will likely endorse a sense of connectedness to other individuals, which may for example be prevalent during lunch breaks with colleagues, but could also reflect a situation in which individuals cooperate in small, face-to-face work dyads. In the following paragraphs I will describe the role of these three identities in a diverse work context. First, I will elaborate on the underlying mechanisms of how subgroup forming and the associated negative effects on social integration outcomes may arise. Second, I will discuss strategies how to reduce potentially negative outcomes in diverse work groups, relying on those three modes of identification. The existing literature has mainly focused on attempting to shift from subgroup identities to more inclusive collective identities or decategorized, personalized identities. I will discuss the role of relational identities as a viable alternative strategy to buffer against negative social integration outcomes in diverse work groups. Subgroup identities in diverse workgroups As I argued previously, generally, perceptions of dissimilarity in demographic attributes are negatively associated with social integration outcomes. These negative effects have predominantly been approached from a self categorization perspective (e.g., Van Knippenberg et al., 2004). Self Categorization Theory predicts that we tend to categorize our social environment into us and them (Tajfel, 1982; Tajfel & Turner, 1979). In this regard, similarity is an important basis for categorization (see Abrams & Hogg, 1990). As individuals have the seemingly universal human tendency to respond positively to similarity and negatively to dissimilarity (e.g., Byrne, 1999), we are attracted by, and feel ourselves associated with, people who have similar attitudes, because they confirm our norms and values and facilitate communication (similarity-attraction hypothesis, cf. Newcomb, 1956). As a result, others are perceived either as members of the same category as the self (ingroup) or as members of a different category (outgroup). Individuals tend to trust and identify themselves more with the ingroup than with the outgroup because sharing the same characteristics reduces uncertainty in social interactions (Hogg & Abrams, 1990). Outgroup members are more likely to be perceived as less trustworthy, honest or cooperative than members of the own ingroup (Brewer, 1979).

12 General introduction 11 Being a member of a diverse work group will likely trigger awareness of differences between members, especially when individuals vary on visible demographic characteristics (e.g. ethnicity or gender). If the subgroup identity is strongly emphasized (e.g. being a white male), subgroup membership will likely be used as a frame of reference and other team members will be categorized in terms of belonging to either the ingroup or the outgroup (cf. Lau & Murnighan, 1998). Resultingly, when team members primarily stress their membership in a subcategory, the emphasis in interactions will be on category values and perspectives, which may differ for the other subgroups within the team. Under those circumstances, incompatibility of category specific values may harm constructive group processes and harmony in the team (Messick & Mackie, 1989). In sum, from a Self Categorization Theory perspective, it can be predicted that increasing diversity makes it more difficult to maintain a sense of we-ness among workgroup members, resulting in the likelihood of subgroup forming, lack of identification, conflict and lower well-being. Interventions aimed at reducing the salience of subgroup identities Previous research has distinguished a number of potential ways to decrease the salience of subgroups in a diverse work group (Brewer & Brown, 1998). These strategies are generally adapted from intergroup interventions theories from the (fundamental) social psychology literature and are largely focused on intervening on the level of collective and personal identities. In the following paragraphs, I will discuss the three major strategies that are described in the literature as useful to reduce intersubgroup tensions within a diverse work group. The recategorization approach assumes that intergroup tensions can be resolved by replacing subgroup identities by a common ingroup identity (Gaertner et al., 1990), thus changing the nature of categorical representation from us and them to a more inclusive we. Recategorization interventions often involve introducing superordinate goals that promote cooperative interdependence among group members. The basic idea is that if members of a diverse work group are stimulated to conceive of themselves more as a single, superordinate group rather than as separate subgroups, team members are more inclined to interpret the world and their own place in it in a manner that is consistent with its values, ideology and culture (Kramer, Brewer, & Hanna, 1996; Mael & Ashforth, 1992), and become more strongly focused and willing to put effort in achieving shared goals. Furthermore, creating a superordinate team identity can reduce negative attitudes and may facilitate more

13 12 Chapter 1 positive affective responses towards other team members (Gaertner & Dovidio 2000; Cunningham, 2005). However, a strong focus on an overarching team identity may have its downsides. According to Brewer (1991), individuals enter social groups with a need to belong and a need to be distinctive. That is, they strive for a sense of belongingness to and unitedness with important social groups while at the same time searching for ways of being distinctive and unique. Whereas the need to belong may be satisfied by identification with groups that are highly inclusive, such as the work team, the need to be distinct may be primarily served by identification with less inclusive minority groups within the work team. In this regard, Optimal Distinctiveness Theory (Brewer, 1991) predicts that group loyalty will be strongest when groups provide its group members a balance between both needs. Some researchers have therefore argued that in the context of diversity the positive impact of team identification on affective outcomes is strengthened when group members are able to combine identification with an overarching category with identification with self-relevant dimensions such as ethnicity, gender and age (e.g. Gaertner et al., 1994; Hornsey & Hogg, 2000). The advantage of combining the two identities is that individuals become full members of a self-relevant overarching category, without having to abandon identification with a self-relevant subgroup; thus creating a balance between a sense of belongingness to and unitedness with important social groups while at the same time searching for ways of being distinctive and unique. Patterns of identification with both the team and self relevant subgroups are usually referred to as dual identities. Empirical findings support the positive outcomes of a dual identity in terms of more harmonious intergroup relationships in an intergroup context (Gaertner et al., 1994). It must be noted that more recent research has shown that intergroup bias and conflict may not be reduced by a dual identity intervention in every intergroup context. Crisp, Walsh & Hewstone (2006) showed that individuals who identified themselves strongly with the own ingroup were less inclined to accept the superordinate team identity and even showed an increase in intergroup bias. In sum, both approaches seem to have advantages in reducing intrateam tensions, but do not seem to be fully successful in every inter(sub)group context. As an alternative to the previous described interventions, in which existing subgroup identities in the team are rearranged in more inclusive collective identities, another approach recommends replacing social identities with identities in which the uniqueness of individuals is central. The essential idea of this decategorization strategy is that intergroup contact will be more effective if interactions are person based rather than category based

14 General introduction 13 (Brewer & Miller, 1984). This implies that individual characteristics rather than characteristics associated with the social categories to which they belong, are used as a basis when individuals engage each other in social interactions. This individuating of the self and others will lead to lower ingroup-outgroup distinctions, as social categories will be less salient. However, decategorization carries the threat of being deprived from valued and desired social identities and may not meet human needs of inclusion and uncertainty reduction (Brewer, 2001; Hogg & Abrams, 1993). Therefore, this strategy may not always be effective as team members are likely to lack the feeling of belonging to a cohesive entity, which may lower the effectiveness on fulfilling tasks in which team members are supposed to cooperate with each other. Managing diversity by focusing on relational identities As an alternative to approaches in which individuals are strongly individuated or in which identification with a more inclusive collective or dual identity is stimulated, it has also been suggested that positive outcomes in diverse organizations may be expected if team members are able to define themselves in terms of their interpersonal bonds with other individuals (e.g., Brewer & Gardner, 1996; Brickson, 2000). In the previous outlined extended model of the self, Brewer and Gardner (1996) expand SIT s notions on personal and collective identities with a self view based on an individual s conception of his/her relatedness to other individuals. Relationally oriented individuals include representations of (close) relationships with others into their mental representations of themselves (cf. Aron, Aron & Smollan, 1992). Positive feelings about the self will derive from developing and maintaining relationships with other individuals (Cross, Bacon & Morris, 2000). Brickson (2000) argued that mutual cognitive understanding and concern among members of different subgroups are most likely to arise from a relational identity orientation. Such an orientation is likely to emerge from dyadic relationships for example organised in small face to face groups or dense and integrated networks. The situation is framed in terms of mutual relations rather than in terms of group membership and the norms that go along with it. Referring back to example of the division within the Democratic Party during the preelections, both Obama and Clinton called upon their friendship in speeches after the elections to re-establish positive bonds between them, and both stated that they were focused on joining hands together to reunite the party, despite their different opinions on certain

15 14 Chapter 1 matters. Thus, the reunion of the Democratic Party seemed to be partly dependent on both candidates tendency to frame the situation in terms of interpersonal bonds rather than in terms of personal interest or interests of the subgroups they represented. The advantage of a relational identity orientation over the previously described interventions based on promoting collective identities seems to lie in the fact that individuals are no longer regarded as depersonalized exemplars of social groups. In addition, in a relational identity orientation context, individuals belongingness needs are better preserved in contrast to the decategorization approach which reflects strong individuating of individuals. This implies that a potential threat to exclusion of the self from the collective or the distinctiveness of the self or will likely be less prevalent when individuals frame the situation in terms of interpersonal relationships. Several studies have indicated that a relational identity orientation promotes the extension of empathy and positive affect, even beyond the interactants to other targets and perceivers (Brickson, 1998, Pettigrew, 1997). Although the positive influence of a relational identity orientation on outcomes in diverse workgroups has theoretically been pointed out (see Brickson, 2000; Brickson & Brewer, 2001), empirical evidence supporting this claim is still scarce (for an exception, see Brickson, 1998). Building on the extended model of the self model of Brewer and Gardner (1996), the present dissertation s central question concerns whether a relational identity orientation can undermine potential negative social integration outcomes in diverse work groups. Context and personality based identities SIT (Tajfel & Turner, 1979) and the extended model of the self (Brewer and Gardner, 1996) recognize that an individual s identity is multifaceted: different levels of identities may coexist within the same individual. Dependent on the context, one level of the self may be activated and dominant over the other levels of the self. On the other hand, cross-cultural psychologists approach an individual s identity from a slightly different perspective by assuming that how an individual s identity is defined, varies between cultures (e.g. Triandis, 1989; Markus & Kitayama, 1991). Members of western cultures tend to think of themselves as independent of relationships and as autonomous or separated from others while members of more collectivistic cultures tend to think of themselves as interdependent with close others as defined by important roles.

16 General introduction 15 Recent work has indicated that cultures are more complex than the basic individualism/independent self collectivism/interdependent self distinction suggests. Therefore, several researchers have argued that there also may be variations within cultures with respect to how the self is defined (e.g. Matsumoto, Kudoh & Tekeuchi, 1996; Cross et al., 2000). In this regard, the dominant perspective in western societies of the self as independent and separate from others has been challenged (see Cross et al., 2000). These authors argue that women in western societies have a strong focus on interpersonal relationships and thus are more likely to view themselves as being relationally oriented rather than being distinct from others. Even more interestingly, individual differences within the sexes with regard to relational interdependent self views have also been acknowledged (Cross et al., 2000). Some individuals are more inclined than others to perceive their social context in terms of interpersonal relationships. Applied to a work context, individual differences in identity orientation may shape individuals perceptions, tendencies and behavior differently in the same situation. For example, entering a diverse work team may for collectively oriented individuals immediately trigger categorization processes and differentiations will be made between team members that are similar and dissimilar to the self. For personally oriented individuals, fellow work team member will be contrasted to the self, which in turn enhances competition tendencies. Finally, for relationally oriented individuals this situation may be perceived as an opportunity to establish interpersonal bonds with other individuals in the team. Taken together, identity orientations seem to be defined and shaped by both the context and by individual differences. The present thesis aims to combine both contextual and individual approaches to the construct of identity orientation to predict social integration outcomes in a diverse work context. More specifically, both personality based relational identity orientation and a relational identity orientation triggered by the situation and the interplay between these two are being examined as a strategy to buffer negative social integration outcomes in diverse work groups. Overview of the chapters The present dissertation focused on the effects of identification patterns on social integration outcomes in diverse work groups. Both context based identities and individual differences in identity orientation are examined with regard to prosocial tendencies towards fellow work group members and well-being in diverse work groups.

17 16 Chapter 1 In chapter 2, I will describe the development and validation of an individual differences measure of identity orientation. Up till now, attempts to construct scales that distinguish between individual, relational, and collective self representations have not yielded clear results with regard to reasonable independent operationalizations of the three different self foci. The present thesis aims to develop a new instrument for measuring these three identity orientations. Several samples are examined to establish the psychometric properties and factor structure of the Identity Orientation Scale (IOS) and its relation with theoretically related constructs. In chapter 3, I will link identity orientation with social integration outcomes in diverse work groups. More specifically, I examine whether a relational identity orientation can promote prosocial tendencies towards members from other subgroups within the same workgroup. In Study 3.1, the focus is on how individual differences in relational identity orientation, measured by the IOS, may affect cooperation tendencies in culturally diverse workgroups. In Study 3.2, outcomes of the contextual salience of different identity orientations on the willingness to help own or other subgroup members will be examined. A priming method is used to make specific identity orientations salient. Effects of the salience of a relational identity orientation on helping will be contrasted with the effects of a personal and a collective identity orientation. Finally, in Study 3.3, I will examine the combined effects of both personality-based and situation-based identity orientations. More specifically, the focus will be on the moderating influence of individual differences in relational identity orientation on the relationship between identity orientation primes and cooperation tendencies towards fellow work group members. Finally, Chapter 4 aims to provide insight into the role of a relational identity orientation as an alternative to contemporary strategies to reduce negative social integration outcomes in diverse work groups due to group members strong identification with self relevant dimensions such as gender, ethnicity and age. As previously was described, interventions aimed at reducing negative effects of this subgroup forming mostly focused on shifting the focus from subgroup identities to an overarching team identity or a combination of both a team and subgroup identity (dual identity). The central aim of this chapter was to examine whether a relational identity orientation could preserve well-being in gender diverse work groups, under conditions of a strong subgroup identity salience. To test this proposition I conducted a field experiment, in which male and female participants worked together on a task and a gender identity, superordinate team identity or dual identity was made salient.

18 General introduction 17 Together, these empirical chapters aim to reveal the important role of identity orientation for diversity research. The central proposition of this thesis is that a relational identity orientation can buffer against the negative effects of diversity on social integration outcomes. Hence, in chapter 5, I will discuss the general conclusions with regard to this proposition, and its implications for research on the self and diversity management. It should be noted in advance that the empirical chapters of this dissertation are based on research that was conducted with several others. Therefore, the term we (instead of I ) is being used in the empirical chapters of this dissertation when any reference is being made to the authors. In addition, the chapters of this dissertation are submitted or under revision as independent articles. As a result, mainly in the introduction sections, there is some overlap between the chapters.

19

20 Chapter 2 Individual differences in self representation: The development of a measure of personal, relational and collective identity orientation 1 The concept of the self is central to an individual s perceptions, evaluations and behaviors in a social context (e.g. Markus & Kitayama, 1991; Triandis, 1989). How individuals define themselves influences how they think, feel, and interact with others. An important notion of an individual s self conception is that it can be construed in multiple ways. Self construal is conceptualized as a constellation of thoughts, feelings and actions concerning one s relationship to others, and self as distinct from others (Triandis, 1994). Up till now, the literature on self representation has been predominantly focused on two levels of self definition: an individual or personal self (those aspects of the self-concept that differentiate the self from others) and a social self (those aspects of the self-concept that reflect connectedness to others or significant social groups; e.g., Markus & Kitayama, 1991; Singelis; 1994; Trifimow, Triandis & Goto, 1991; Triandis 1989). However, when looking at the social self, a further distinction between two levels of social self seems feasible to make: a level of self that derives from interpersonal relationships with specific others, i.e. the relational self, and a level of self that derives from membership in larger, more impersonal collectives or social categories, i.e. the collective self (Brewer & Gardner, 1996; Cross, Bacon & Morris, 2000). Research on the different levels of the self mainly originates from two different theoretical research traditions. On the one hand, context based theories, such as social identity theory (Tajfel & Turner, 1979) and social categorization theory (Turner et al., 1987), assume that different levels of the self coexist within the same individual. Dependent on the context, one level of the self may be activated and dominant over the other levels of the self. On the other hand, cross-cultural and (social) personality psychologists approach the self from a slightly different perspective, assuming that self representations are more or less stable internalized in the self concept. How an individual s self is construed, varies between cultures (e.g. Triandis, 1989; Markus & Kitayama, 1991) or between individuals within the same culture (e.g., Cross et al., 2000; Cross & Madson, 1997). Up till now, existing personality approaches to self representation are mainly focused on operationalizing the classical distinction between the personal/individualistic and social/collective self. Recently, a few authors have extended this two way distinction and focused on elaborating constructs referring to the relational self (Brewer & Gardner, 1996; Cross et al., 2000). The present study aims to integrate this extended model of the self with an individual difference perspective on self representation. More specifically, we will describe the development and validation of an 1 This chapter is based on Vos, Van der Zee & Buunk (2009).

21 20 Chapter 2 instrument aimed at measuring individual differences in the three levels of self representation, namely the personal self, the relational self and the collective self (Identity Orientation Scale IOS). First we will discuss the concepts of the three self representations more in-depth. Second, we will present two studies in which our IOS scale will be validated and linked to theoretically related psychological constructs. Conceptual approaches to self representation The notion that self representation is multifaceted has early been acknowledged by social identity theorists (Tajfel & Turner 1979). Social identity theory assumes that people enter social situations with a self-concept that is basically determined by their personal identity and their social identity (see Turner, 1982). Personal identity is the individuated self those characteristics that differentiate one individual from others within a given social context. Social identities are categorizations of the self into more inclusive social units that depersonalize the self-concept (Tajfel & Turner, 1986). When social identities are salient, individuals have a tendency to categorize themselves as members of social groups and the characteristics that belong to the group will be internalized in the self. Brewer and Gardner (1996) recognized that the social self does not always have to be based on group identities, but can also revolve around interpersonal bonds with other individuals. In their extended model of the self, they proposed a theoretical framework in which three distinct self views are central: the personal (individual), relational (interpersonal) and collective (group) identity orientation. The personal identity orientation refers to an individual s self conception primarily in terms of their individual traits and characteristics. A positive view about the self is derived from a sense of uniqueness compared to others. These individuals have therefore the tendency to contrast themselves from others (see Stapel & Van der Zee, 2006). The relational identity orientation refers to an individual s conception of their relatedness to other individuals. In other words, they include representations of (close) relationships with others into their mental representations of themselves (cf. Aron, Aron & Smollan, 1992). Positive feelings about the self will derive from developing and maintaining close relationships with other individuals (Cross et al., 2000). Examples of interpersonal identities are those derived from intimate dyadic relationships such as friendships but also networks of dyadic relationships such as small, face-to-face work groups (Brickson, 2000). Finally, the collective identity orientation revolves around a self conception in terms of being a group member and describing oneself in terms of characteristics that are connected to the group. Furthermore, these individuals are likely to emphasize group affiliation, ingroup

22 Individual differences in self representation 21 norms and roles and status defined by collectives (Kashima & Hardie, 2000; Tajfel & Turner, 1979). Collective identities, as opposed to relational identities, do not necessarily require interpersonal relationships among group members. Although Brewer and Gardner acknowledge that there may be individual differences in identity orientation, their model primarily focuses on contextual properties of self representation: different self representations coexist in the same individual and can be activated at different times and in different contexts. Alternatively, cross-cultural and (social) personality psychologists approach the self from a slightly different perspective, assuming that self representations are more or less stable internalized in the self concept. The crosscultural psychology literature assumes that how the self is construed, is imbedded in the specific culture (e.g. Markus & Kitayama, 1991; Triandis, 1994). In this regard, members of western cultures tend to think of themselves as independent of relationships and as autonomous or separated from others. The primary components of this independent self view are one s unique traits, abilities, preferences, interests, goals and experiences, and these are differentiated from social contexts, interpersonal relationships, and group memberships. By contrast, members of more collectivistic cultures tend to think of themselves as interdependent with close others and as defined by important roles and situations. The underlying principle that determines these interdependent self views is the notion that individuals are connected to others, so that the self is defined by important roles, group memberships and interpersonal relationships with others. Cross-cultural research on the independent and interdependent self has studied effects of different self representations on ratings of expression of emotions (Matsumoto, Kudoh, Scherer & Wallbott, 1988), coping with stress (Cross, 1995) and reactions to threat to the self (Brockner & Chen, 1996). Recent work has indicated that cultures are more complex than the basic individualism/independent self collectivism/interdependent self distinction suggests. Therefore, several researchers have stated that there also may be variations within cultures with respect to how the self is defined (e.g. Greenfield, Keller, Fuligni & Maynard, 2003; Matsumoto, Kudoh & Tekeuchi, 1996; Cross et al., 2000). Greenfield et al. (2003) argue that independent and interdependent concerns are prevalent in the same culture. More specifically, an individual s developmental pathway across their lifespan is characterized by a focus on either individuation and independence or group membership and interdependence. For example, interdependence is strongly emphasized during the relationship formation phase at birth and early childhood with other family members, whereas adolescence and early adulthood is characterized by a strong focus on individual choice and independence. In addition, studies on gender differences challenge the dominant perspective in western societies of the self as independent and separate from others as well. Cross and

23 22 Chapter 2 Madson (1997) argued that women in western societies have a strong focus on interpersonal relationships and thus are more likely to view themselves as connected to or interdependent with others rather than as individuals separated from others. In contrast, Wainryb & Turiel (1994) found a stronger orientation toward autonomy and independence among men than women. There also appears to be variation in how the self is construed within the sexes (e.g. Cross et al., 2000; Kashima & Hardie, 2000). According to Higgins (1990), individuals differ in terms of which elements of self knowledge are frequently activated. This may correspond to a predisposition among individuals towards a particular identity orientation. Some important issues have been raised with regard to the individualism/ independence collectivism/interdependence framework. One aspect of these criticisms addresses the questionable empirical support of the theoretical perspective. In this regard, Kagitcibasi (2005) pointed out that many measures of individualistic/independent self views confound agency (autonomy-heteronomy) with interpersonal distance (separatenessrelatedness). In addition, the (cross-cultural) literature s more collectivistic group-oriented notion of interdependence, in which the self is defined in terms of the position in the group, does not adequately seem to fit the relationship centered conception of interdependence that characterizes western cultures (Cross et al., 2000; Kashima et al., 1995; Kashima & Hardie, 2000). Most existing scales of interdependent self contain elements referring both to connectedness to other individuals as well as connectedness to social groups. We argue, in line with Brewer and Gardner s extended model of the self, Cross et al. (2000) and Kashima et al., (1995), that the concept of interdependent self can be subdivided into a relational self view component, which stresses relationships and connectedness with other individuals, and a collective self view component, which refers to connectedness with social groups and regarding oneself a group member. Although earlier attempts have been made to construct scales that tap into the differences between these conceptualizations (e.g. Kashima & Hardie, 2000), reasonable independent operationalizations of these individual differences in the three different self representations have not yielded clear results. In the present research, we aimed at developing a valid instrument for identity orientations that clearly discriminates between the personal, relational and collective self. Overview present research The aim of the present study was to develop and evaluate the psychometric properties of an instrument aimed at measuring individual differences in the three levels of self representation (Identity Orientation Scale IOS). In terms of psychometric properties evaluating, we focused on the reliability and construct validity of this instrument. The construct validity of

24 Individual differences in self representation 23 the IOS was examined by relating the scale to theoretically related psychological constructs. In study 1, we will describe the development of the IOS, its psychometric properties, and its relations to other personality constructs. Furthermore, gender differences in identity orientation will be discussed and analyzed. In study 2, construct validity will be further examined by relating our measure with other self construal concepts. In addition, we will test stability over time to further establish the reliability of the identity orientation scale. Again, gender differences will be examined. Study 1 In Study 1 we report the development ad psychometric properties of the IOS. Construct validity was assessed by relating the IOS with the Big Five personality traits (Goldberg; 1993; Hendriks, Hofstee & de Raad, 1999), and closely related constructs, i.e. communal orientation (Clark, Ouellette, Powell & Milberg, 1987; Buunk, Doosje, Jans & Hopstaken, 1993), collective self esteem (Luhtanen & Crocker, 1998) and measures of independent/ interdependent self construal (Singelis, 1994). Furthermore, we were interested in gender differences in identity orientation. First, we expected that the interpersonal traits of the Big Five, namely extraversion and agreeableness, were particularly related to the relational identity component of the IOS. Extraversion, the first factor of the Big Five factors of personality refers to an individual s tendency to be sociable, to disclose oneself easily to seek high levels of activity (Hendriks et al., 1999). A focus on interactivity with others seems to tap into both relational and collective self conceptions, as these orientations reflect the social spectrum of the self. However, self disclosure seems particularly relevant in interpersonal relationships as the sharing of personal information seems to strengthen bonds between individuals (e.g. Buhrmester, Furman, Wittenberg & Reis, 1988). For collectively oriented individuals self disclosure will be less relevant because their focus will be not as much on improving bonds with other individuals as relationally oriented individuals, but more on fitting into the group or adapting to collective goals. We therefore expected extraversion to be particularly related to our measure of relational identity orientation. Agreeableness, the second factor of the Big Five factors of personality, reflects individual differences in concern with cooperation and social harmony. Agreeable individuals value getting along with others and are considerate, friendly, generous, helpful, and willing to compromise their interests with others. Although the tendency to strive for social harmony may be relevant in collectives, we expected that agreeableness will particularly be related to a relational identity orientation. Collective identities often involve

25 24 Chapter 2 depersonalized conceptions about the groups to which one belongs and therefore do not necessarily require interpersonal relationships among its members. Consequently, concern for the group and its goals may be more relevant for these individuals than concern for other individuals. A relational identity orientation, on the other hand, involves a strong focus on maintaining harmonious relationships and a high concern for others. We therefore expected agreeableness to be particularly related to our measure of relational identity orientation. In contrast to these two social traits of the Big Five, we expected autonomy, which refers to an individual s tendency to take independent decisions without being influenced by the social environment, will particularly be related to our measure of personal identity orientation, as both concepts share the notion of the self being independent of others. We have no specific predictions about the relatedness of IOS with the remaining Big Five personality traits i.e. conscientiousness and emotional stability - as these traits do not, or to a lesser extent, directly involve a social environment. Second, we expected that communal orientation will be strongly related to the relational identity orientation. Communal orientation refers to a trait that encompasses the desire to give and receive benefits in response to the needs of and out of concern for others (Buunk et al., 1993). Clark et al. (1987) show that individuals high in communal orientation have a tendency to help other individuals more than individuals low in communal orientation. As the tendency to procure the welfare of other is associated with the relational self representation (Brewer & Gardner, 1996), we expected that communal orientation will particularly be related to our measure of relational identity orientation. Third, we expected collective self esteem to be particularly related to the collective identity orientation. Collective self esteem reflects the positive feelings that individuals derive from their social group and the values they place on their social group (Luhtanen & Crocker, 1998). Feeling positive about one s group can be attained by favorable comparisons with relevant other groups (Tajfel & Turner, 1986), which is also a central notion in the concept of the collective identity orientation. Because collectively oriented individuals derive their self esteem from their memberships of social groups (Brewer & Gardner, 1996), we expected that our measure of collective identity orientation will be positively related to collective self esteem. Fourth, we expected our measure of identity orientation to be related to existing measures of independent and interdependent self (Singelis, 1994). More specifically, independent self will be related to our measure of personal identity orientation and the interdependent self measure will be related to both relational and collective identity orientation. Finally, we were interested in gender differences in self representations. Research on gender roles and stereotypes (e.g. Broverman et al., 1972; Eagly & Wood, 1999; Glick &

26 Individual differences in self representation 25 Fiske, 2001; Williams & Best, 1982), showed that men are generally perceived as independent and competitive towards others and women as relationship oriented and cooperative towards others. Therefore, Cross & Madson (1997) proposed that men have independent self schemas, whereas woman mainly have interdependent self schemas and these differences can be explained on the basis that men prefer separateness and independence, whereas women generally seek to form intimate connections with other individuals. These preferences seem to be shaped in early childhood when norms of autonomy and competition are adopted by boys and norms of cooperation and efforts to maintain social relationships by girls (see Maccoby, 1990). Based on these findings, we expected men to be more likely to define themselves in terms of their uniqueness compared to others and women to be more likely to define themselves in terms of close relationships with others. In other words, we hypothesized that men are more personally oriented compared to women, whereas women would be more relationally oriented compared to men. Method Sample and procedure 376 first year psychology students participated in the first study during a mass testing session. The sample consisted of 98 men (26.7 %) and 278 women (73.3 %). The mean age was 20.3 (SD = 2.1) years old. All participants filled in a questionnaire, which contained the measures described below. Measures Identity Orientation Scale (IOS). Based on the literature (e.g. Brewer & Gardner, 1996; Triandis, 1994; Cross et al., 2000) and the conceptualizations of the three levels of the self given in the introduction, items for the IOS were generated by the authors. Items referring to self conception in terms of individual traits and characteristics and the sense of uniqueness compared to others were developed for the personal identity orientation scale. Sample items are: I enjoy being different from others and It is important for me to do my own thing. The items for the relational identity orientation scale referred to a self conception in terms of relatedness to other individuals, i.e. individuals representations of (close) relationships with others into their mental representations and their focus on maintaining relationships with others. Sample items are: I enjoy maintaining personal relationships with others and I think that close others have much influence on my identity. Finally, items referring to self conception in terms of group memberships and the connectedness to the groups to which one belongs were developed for the collective identity orientation scale. Sample items are: I