Running Head: Cognitive and Affective Identification
Cognitive and Affective Identification: Exploring the Links between Different Forms of Social Identification and Personality with Work Attitudes and Behavior
Michael D. Johnson
University of Washington
Frederick P. Morgeson
Michigan State University
David R. Hekman
University of Wisconsin-Milwaukee
Under 2nd review at the Journal of Organizational Behavior
Individuals often identify with groups in order to either reduce perceived uncertainty or to feel better about who they are as individuals. This suggests that cognitive and affective identification are two distinctive forms of social identification in organizational settings. Because neurotic individuals are highly motivated to reduce perceived uncertainty, they will tend to identify cognitively with groups. Extraverted individuals, on the other hand, are highly motivated to enhance how they feel about themselves and thus identify affectively with groups. Across three studies, we develop measures of cognitive and affective identification and then show that neuroticism is positively related to cognitive identification, whereas extraversion is positively related to affective identification. We also find that affective identification provides incremental predictive validity over and above cognitive identification in the prediction of organizational commitment, organizational involvement, and organizational citizenship behaviors.
Key words: Social identity, scale development, personality, cognitive ability, cognition, affect, emotions
Social identification is the sense of oneness that group members feel with certain groups, and the degree to which they define themselves as members of those groups (Ashforth & Mael, 1989). When employees identify with organizational groups (i.e., workgroup, team or organization), they are less likely to leave (Ashforth & Saks, 1996; van Knippenberg & van Schie, 2000), perform more organizational citizenship behaviors (Bartel, 2001; Bergami & Bagozzi, 2000), are more involved on the job (van Knippenberg & van Schie, 2000), are more satisfied with their job (Ashforth & Saks, 1996; Lee, 1971; Mael & Ashforth, 1992; Mael & Tetrick, 1992), and report working harder (Ashforth & Saks, 1996). In short, social identification appears to lead employees to exert more effort on behalf of the groups they identify with.
From its initial conceptualization, social identification was thought to include both cognitive and affective dimensions. Tajfel (1972) defined social identity as involving an individual’s knowledge of group membership and the emotional significance the individual attaches to that membership. Groups provide a way for individuals to place themselves and others in society such that individuals cognitively define themselves as organization members (Albert et al., 1998). Groups also provide a sense of pride in the group and belongingness, and reflect the emotional value of that group to the group member (Albert et al., 1998). The theoretical underpinnings of social identification are based on a hybrid of two streams of research: self-categorization theory (which is largely the basis of the cognitive aspect of identification) and social identity theory (which is largely the basis of the emotional aspect of identification).
Drawing on these rich theoretical streams, we suggest there are two main reasons why people identify with groups: (1) to feel better about themselves (the “self-esteem hypothesis;” Tajfel & Turner, 1979) and (2) to reduce social uncertainty (the “uncertainty-reduction hypothesis;” Hogg, 2000; Reid & Hogg, 2005). We propose that self-esteem is associated primarily with affective identification, as it relates to how one feels about oneself and the group. In contrast, uncertainty-reduction is associated with cognitive identification, as it relates to how one thinks about oneself and defines one’s place in the social environment.
Moreover, we draw on personality theory (Costa & McCrae, 1992) to hypothesize that two traits are associated with these different motives for identifying with social groups. In the past, social identification researchers have tended to focus solely on the situational determinants of identification, and have neglected potential dispositional antecedents (see Riketta  for a review). This lack of attention to individual differences in the prediction of identification is unfortunate because identification researchers have suggested that people may differ in their “propensity to identify” (Albert et al., 1998, p. 238). Some individuals may simply be more likely than others to identify with any social group. We outline how extraversion and neuroticism are two personality traits that are likely to relate to the two motives for identifying with social groups, and thus with the two identification dimensions.
Because identification research has not focused enough attention on the sources and outcomes of cognitive and affective identification, we seek to identify the dimensions of identification and empirically test their antecedents and outcomes. In so doing, we develop a rigorous two-factor measure of identification that distinguishes affective identification from cognitive identification, examine the dispositional sources of these two forms of identification, and test the incremental validity of affective identification over cognitive identification.
Two Dimensions of Social Identification
In an attempt to distinguish identification from commitment, organizational researchers may have focused too heavily on the cognitive, self-definitional aspect of social identification. The primary difference between identification and commitment, however, is not a cognitive vs. affective one. Instead, it is that identification reflects the sense of oneness with a group, whereas commitment reflects the relationship strength between separate psychological entities (Ashforth and Mael, 1989; Meyer & Allen, 1991; Pratt, 1998; van Knippenberg & Sleebos, 2006)a. A similarity between the two constructs is that there is consensus that both identification and commitment have cognitive and affective components (Cheney, 1983; Edwards & Peccei, 2007; Harris & Cameron, 2005; Meyer & Allen, 1991; van Dick et al, 2004). Although the affective component of commitment has been extensively examined, the affective component of identification has been the subject of very little empirical research.
Ashforth and Mael (1989, p. 21) acknowledged that their concept of organizational identification deviated from existing social identification research because it excluded the “affective and evaluative” dimension of identification. More recently, organizational researchers have suggested that research should re-examine the unmeasured affective dimension (Abrams, Ando and Hinkle, 1998; Albert et al., 1998; Ashforth, Harrison, & Corley, 2008; Smidts, Pruyn, & van Riel, 2001). The lack of attention to the affective dimension of social identification may be responsible for the surprising lack of support for one of the main reasons why people are thought to identify with groups – to make themselves feel better. Indeed, this “self-esteem hypothesis” (Tajfel & Turner, 1979) is a core tenet of social identity theory and suggests that because individuals prefer a positive rather than negative self-image, they will identify with groups that enhance their sense of self-esteem (Reid & Hogg, 2005). However, this hypothesis has received only mixed support, as Rubin and Hewstone (1998) found that only half of studies confirmed the self-enhancement motive of social identification (see also Crocker & Luhtanen, 1990; Long & Spears, 1998).
We suggest that this hypothesis has received mixed support in part because the most common measures of social identification do not capture how people feel about their group membership, making the link between self-esteem and social identification difficult to detect. Existing identification measures still focus almost exclusively on the cognitive aspect of the construct (Harris & Cameron, 2005; Cheney, 1983) – individuals’ self-perceptions that they are one with a group (Dutton, Dukerich & Harquail, 1994; Ashforth & Mael, 1989). Indeed, the most popular social identification measures ask people whether they define themselves as group members, not how they feel about being part of the group (e.g., Mael & Ashforth, 1992). Certainly, the first and most basic dimension of identification is cognitively categorizing oneself as a group member (Ashmore Deaux, & McLaughlin-Volpe, 2004). This is generally recognized as the heart of social identification (Deaux, 1996). Self-categorization, or what we call cognitive identification, may be the precondition for someone to feel any type of emotions related to their identification. For example, to feel proud of being a member of a particular group, individuals may first cognitively place themselves into this category (Phinney, 1995).
What we label affective identification is how people feel about themselves in relation to a particular social group (Ashmore Deaux, & McLaughlin-Volpe, 2004). We suggest that affective identification reflects individuals’ feelings of oneness with the group. Feeling oneness with the group is distinct from perceiving oneness with the group (e.g., existing measures of organizational identification), and should involve positive feelings about one’s membership, including pride and happiness (Albert et al., 1998). Whetten (in Albert, 1998, p. 11) likens identities to onions where, as one peels back the layers, one eventually elicits “tears” (strong emotions), signifying the vital core of identity that is the basis for identification. Indeed, the “tears” can reveal to oneself and others what one in fact identifies with. In short, as Harquail argues (in Albert et al., 1998, p. 225), “identification engages more than our cognitive self-categorization and our brains, it engages our hearts.” Accordingly, we define affective identification as an individual’s positive feelings about being one with a group.
We suggest that the affective dimension of identification that is actually experienced is generally positive (e.g., pride, excitement, joy, love) because individuals who can say “I am ‘A’ and it’s important to me,” wish to feel positively about their membership and often find sources of pride in even the most stigmatized of groups (Ashforth, Harrison & Corley, 2008; Cameron, 2004). Individuals may therefore act to make the group an ongoing source of positive affect, and thus may generally feel positive about it. Further, individuals who generally feel positively about the group are more likely to conclude, “I am ‘A’ and it’s important to me.” Thus, cognition and affect reciprocally reinforce identification (e.g., Kessler & Hollbach, 2005). Indeed, Pratt and Ashforth (2003, p. 313) suggest that “work, over time, tends to implicate one’s sense of self such that behavior, cognition, and affect converge over time.”
We aim to disentangle social identification by winnowing the dozens of social identification scales into two dimensions – cognitive identification (the self-categorization aspect of identification) and affective identification (the emotional and evaluative aspect of identification). We believe this refined approach to identification will enable greater theoretical precision about where social identification comes from within organizations. Particularly, we are interested in the defining employee personality traits that enable individuals to identify cognitively or affectively with a group.
Two Motives for Social Identification
Although much existing research on social identification has examined its situational antecedents, the degree to which individuals identify with their organizations and other work-based groups may be affected by dispositional differences, such as individuals’ innate need for identification (Kreiner & Ashforth, 2004). We suggest that two personality factors are particularly likely to affect levels of identification: extraversion and neuroticism. In essence, these factors may make an individual more likely to identify with social groups in general, regardless of the dynamics surrounding any particular social group. The reasons they lead to identification differ, however, and thus lead differentially to cognitive and affective identification.
As noted above, social identity and self-categorization theories suggest people identify with groups for two main reasons – to feel better about themselves (the “self-esteem hypothesis;” Tajfel & Turner, 1979) and to reduce painful social uncertainty (the “uncertainty-reduction hypothesis;” Hogg, 2000; Reid & Hogg, 2005). We suggest that these motives cause different aspects of individual’s identities to become salient (Ashforth & Johnson, 2001); specifically, we argue that the dispositional traits of extraversion and neuroticism are associated with the self-esteem motive and uncertainty reduction motive respectively, and in turn are associated with the two dimensions of identification.
Extraversion leads to a desire for self-enhancement (Matthews, Zeidner, & Roberts 2004), which reflects a motivation to emotionally connect with one’s social environment. Extraversion has been shown to be moderately to strongly related to various group process variables, such as cohesion, communication, flexibility, and conflict (Barrick et al., 1998; Barry & Stewart, 1997). Extraverted people enjoy working in groups and like to be in the presence of other people (Costa & McCrae, 1992). This enjoyment appears to make extraversion a likely candidate for predicting affective identification. As noted above, affective identification emphasizes positive feelings of oneness with the group (Tajfel, 1981); extraverted people should, in general, have stronger feelings of oneness with the groups of which they are members because of their preference to be in group situations.
Moreover, extraversion is the personality trait most consistently associated with positive affect (Larsen & Ketelaar, 1991). Indeed, Costa and McCrae (1980) found that extraversion positively predicted positive affect ten years after the personality assessment, and concluded that extraversion “predisposes individuals to positive affect” (p. 673). Affective identification is thought to be associated with positive feelings about one’s membership in a social group (Albert et al., 1998), and by extension, extraverted individuals with their typically higher levels of positive affect are more likely to experience positive emotions about their social groups.