Here is a brief summary of what I observed both in my one published study and the dissertation study I never defended that I will present in chunks here as creative rather than professional writing:
People are willing to discuss key group beliefs (like being pro automatic weapon and Republican) with other individuals or breakout groups that are on the fence only if there is some extenuating circumstance like the individual being new to the group or a breakout group having other things in common. Otherwise, nobody wants to hear ambivalence. For example, if you disagree with a fellow so-and-so (let’s say football player) about student athletes getting paid beyond their scholarships, tell them you are new to student athletics or the debate over whether they should get paid and they may stay your friend. In the context of a small group of people, they might be more willing to work through differences of opinion if they already know they share the same personality types, so a group leader might want to emphasize this similarity, called intra-group similarity to get people actively trying to get on the same page.
Another way to say the second finding is people are more willing, especially when feeling uncertain about themselves and the future, to work through disagreement when the group is similar not in opinions but otherwise, presumably because if the point of working together is cooperation then when uncertain we want to find the quickest path to agreement and only if similar in some other way is there hope of working through disagreement.
What follows is the intro to the dissertation study that gave evidence for this self-evident truth, that people need some reason to talk to you or hear you out if you disagree, either that they’ll be able to convince you they are right because you are otherwise likeminded, that you will eventually change your mind because you’re new to things, or that the thing being debated is not that important, written up in academese:
CHAPTER I
Background & Justification
Maintaining a positive and distinct self-concept (social identity theory, Abrams & Hogg, 2012; Tajfel & Turner, 1986), and reducing uncertainty with regards to the content of this self-concept (uncertainty-identity theory, Hogg, 2007) motivate identification with social groups, including any readily perceived as coherent, well-defined, and distinct entities (i.e., high in entitativity; Lickel, et al., 2000). Self-categorizing into a group allows people to adhere to a hypothetical ingroup prototype, the ‘fuzzy’ set of traits defining a social category (self-categorization theory, Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), thereby reducing uncertainty even in people in artificially created groups (Grieve & Hogg, 1999). Identity-uncertainty, for example, uncertainty about a personally important group’s social standing, can also inversely motivate behavior, like derogation of deviants, by highly identified ingroup members, that serves to uphold the relative clarity of ingroup prototypes (Marques, Abrams, Paez, & Martinez-Toboada, 1998). The tendency of decision-making groups to uphold and accentuate pre-existing choice preference norms in lieu of objective means for making the most informed collective decision also stems from a desire to reach or maintain a valid consensus (Isenberg, 1986; Janis, 1973; Stasser & Titus, 1985; 2003). Furthermore, groups formed in the crucible of uncertainty tend to endorse more polarized or extremist (Hogg, Adelman, & Blagg, 2010; Nash, McGregor, & Prentice, 2011) and risky (Hogg, Siegel, & Hohman, 2012) behaviors that more rigidly enforce intra-group similarity and inter-group differences. This altogether suggests decision-making group members feeling uncertain would identify more with groups with a clear preference norm or easily ascertainable majority preference (especially a preference for more risky behavior if given a choice between risk and caution), in the absence of traits furnishing the perception of a group as high in entitativity (and lack of a skewed preference norm, disagreement or ambivalence, may render such traits as intra-group similarity significant predictors of identification and cohesion between self-uncertain people and their disagreeing or ambivalent group members).
Introduction
In task-oriented groups, similarity (e.g., having the same role, being the same age) can have negative or positive effects on identification, cohesion, and performance depending on, among other things, the nature of the task a group was formed around. van Knippenberg et al. (2011), for example, found that if a task involves brainstorming the highest number of unique responses, and people are placed in groups described as having different cognitive styles, people will appreciate that their group is better equipped to complete this task and identify with it more. Especially in the presence of motivational factors that remind people of their needs for inclusion or individuality, groups can seem either too diverse and hard to define or too similar and stifling (Crisp et al., 2010; Dovidio et al., 1998; Hornsey & Hogg, 2000). Intra-group dissimilarity can decrease cohesion (O’Reilly et al., 1989) and member commitment (Riorden & Shore, 1997), increase group turnover (Wagner, Pfeiffer, & O’Reilly, 1984), and increase relational conflict between group members (Pelled, Eisenhardt, & Xin, 1999). If similarity is perceived, in a context where being unique is not important (e.g., without bringing into question an ingroup’s positive distinctiveness), or similarity is balanced out by there being differences that allow people to maintain some distinctions, people identify more with groups they are (put) in(e.g., Brown & Wade, 1987). In essence, people prefer or identify more with groups that balance inclusion and individuality both within the group, as above, and between groups (Brewer, 1991). In decision-making groups formed around reaching a single collective choice, people can be more-or-less different in several ways that all could lend themselves to uncertainty reduction; the choices each individual prefers, incidental traits unrelated to the choice dilemma, and basic norms of human interaction can be more or less discernible.
Self-uncertainty motivates identification with groups because identification reduces it presumably theoretically through self-categorization processes whereby people take on traits that are presumed to be prototypical of a group and without a clear inkling of what these traits are, a group does not do well to reduce uncertainty. Self-uncertainty should then also motivate identification with groups that can be presumed to be leaning a certain direction opinion-wise based on other markers including those supplanting entitativity perceptions or because they are explicitly described that way. Decision-making groups thus provide a specific applied context where both of these, decision choice preferences and choice-irrelevant traits, can vary or be varied independently in the presence of primed self-uncertainty. Previous research has looked at these two things as both a part of a group’s appearance as a unit, not treating choice preference as something separate from entitativity or the appearance of the group as a unit. The characteristics of a group, like intra-group similarity and inter-group differences, which support the perception of a group as entitative or group-y, are often otherwise correlated, and related to the content and clarity of the group-prototypical position, and some social psychological phenomena observed primarily in decision-making and persuasion research, rest on the expectation that they, normative beliefs, attitudes, or opinions and other incidental traits. Examples being what diversity researchers have termed variety diversity, heterogeneity on one dimension within a group (i.e., dissimilarity), potentially breeding an expectation of conflict or argument in a cooperative group (Jehn & Ayub, 2014), and the salience of categories with self-evident attitudinal norms (e.g., students being expected to prefer there be no undergraduate exit exams) leading to accentuation of these norms especially if they become known (Hogg, Turner, & Davidson, 1990). Perceptions of similarity and clear preference norms can bolster each other when the category that people share through similarity relates to what the average and favored option in a group seeking a unanimous preference is. Plus, just being in an experimental group leads people to follow the general norm of ‘let’s get along and get this task completed’, so agreement is enhanced and presumably preferred for this reason as well (Hogg et al., 1990).
On the other hand, a group or individuals addressed together can also appear entitative or be similar in a way that is independent of what and how clear the content of its norm or prototypical position or opinion is and reduce self-uncertainty of identifying members more than less entitative groups, either leaving people open to diverse opinions or adding to the uncertainty-reductive effect of identifying with a group that is more likely to find agreement (regardless of) if the group actually seems to agree or prefer the same choice. Meaning: the question remains as to whether there is more of an additive or a necessary and sufficient thing going on here where similarity is enough to lead people to identify more with a group and (dis)agreement will have no additional effect, or similarity and agreement together lead to the most identification in self-uncertain people. If a group of people polled about their political values was a mix of different political affiliations, there would be little expectation of a majority opinion leaning one way or another, but if the job, the purpose given to the group, was to find some common ground, the more people were similar in any way or had a clear majority preference, the more at least this group would have a clear path to consensus (reducing task uncertainty that is related to self-uncertainty [Choi & Hogg, 2019]), possibly leading self-uncertain people to identify more with a group that looked like it would find agreement easier. If no single political affiliation and no majority preference existed in the survey group, it is still conceivable that being similar on any trait including height, totally unrelated to how much people agree or the political topic, would make the group seem more a unit than if they shared no traits in addition to simply being in the same polling group, again leading uncertain people to identify more relative to dissimilar groups. Even (dis)similarity in age has been shown to predict stereotyping of task group members and perception of potential task conflict (Tsui, Egan, & O’Reilly, 1992), so similar heights could also trigger an expectation of agreement, though, in which direction people of the same height would be expected to lean on political attitudes it is impossible to say. In this case, being assigned to a group of people the same height might make people feel a part of a more easily defined group, reducing uncertainty via identification, independently of there being a majority opinion. Lastly, people could mostly agree on being unsure what to choose, in which case the group would have an average or normative preference, but not one skewed in any direction, therefore this group would not have a clearly prototypical preference and may disidentify with the group if it is also not similar in any other way, related or not to the choice scenario. All these possibilities are represented in the reported research, the prediction being that one or the other, agreement or similarity, is enough to increase identification in self-uncertain people, implying a reduction in self-uncertainty will occur in either case and not more so when a group agrees and is similar.
Similarity supporting entitativity perceptions and aspects of a group furnishing a clear prototype (agreement), as in when there is a clearly favored option (and the normative preference or average score on a measure of attitudes is not in between two options or in between being pro- or con-) have in the past been manipulated very generally as intra-group similarity and inter-group differences together under the umbrella of entitativity (Hogg et al., 2007), or the extent to which a group of people are perceived as a (social) unit, but are not the same or even both always related to entitativity perceptions. Within the context of decision-making groups, intra-group similarity can be irrelevant to the choice scenario or expected normative attitude in question, but self-uncertain group members may still use similarity as a cue to assess how difficult it is to reach a valid unambiguous consensus, and thus disidentify with groups lacking a polarized norm if they are also dissimilar because neither necessary conditions for uncertainty reduction (a prototype or polarized norm to adhere to or traits implying there is one) are there.
In cases where they are related explicitly, for example when surface-level, apparent differences map onto differences of opinion or task-relevant skills, and the motives (like self-uncertainty) for identification with distinct and defined groups are present, group members may ignore or not work through disagreement and may otherwise expect more conflict and be less satisfied with the group (Jehn & Mannix, 2011). On the other hand, when not explicitly correlated, similarity may reduce uncertainty (where it otherwise might reduce positive distinctiveness) enough to free people up to appreciate especially task-relevant differences (van Knippenberg & Schippers, 2007), but perhaps also appreciate equally favored and opposing preferences also called full disagreement that leads to more sharing of different opinions in decision-making groups (Schulz-Hardt et al., 2006). The idea being that identification under self-uncertainty may track with people’s tendency to address full disagreement (Brodbeck et al., 2002) if there is intra-group similarity to suggest that consensus is a realistic possibility even given disagreement. The assumption of the present research is that self-uncertain people prefer (i.e., identify and want to work more with) groups that in the decision-making literature tend to address task-relevant differences when these differences are balanced out by similarity on task-irrelevant traits that reduce self-uncertainty as evidenced by increased identification.
Social Identity and Identification
Identity or self-concept is the most difficult concept to pin down without using circular definitions because it is not only the content that people use to answer the question of who they are but also the faculties that are brought to bear on retrieving the answer to this question (Markus & Wurf, 1987). Practically speaking, identity is a collection of descriptors that apply to an ‘I’ and suggests how a particular ‘I’ or person is positioned relative to others with the same or different self-ascribed labels. Identity is a schema (Fiske, 1992) for interpreting our social environment; who ‘I’ is determines what offends me, what motivates me, where I fit into different social hierarchies, and what behaviors can be appropriately expected from me depending on social context. How a person interacts with others varies based on the labels appropriately applied to them—the same person called a co-worker or spouse is expected to behave in appropriately different ways in these different contexts. Self or identity can thus be construed or described intra-personally or as one’s cognitive processes or interpersonally, embedded in the context of a dyadic relationship, or as belonging to a social category or group (Brewer & Hewstone, 2004). Social identity is the part of the self-concept defined by our group memberships and insomuch as a group is important to defining a person, that person will strive to maintain a positive and unique image for this group as he or she is wont to do for his or her personal identity.
In the terms of social identity theory (Tajfel & Turner, 1986), social identification happens between an individual and groups the individual consciously feel a member of and where membership has some sort of hedonic relevance or emotional significance. That means that in general or based on the purpose of the group in the moment, people are invested in how well the group does or reflects on the individual. Most measures of social identification include subscales assessing how central to defining the self-concept a group is, as well as how emotionally attached to the group we are, and how much we feel we fit the social category (Cameron, 2004). Though sometimes operationalized as the same thing (Turner & Pratkanis, 1994), cohesion is more a reflection of how much people in a group identify as a whole and how positively they feel toward other members or really how intimate, friendly, and ‘all for one, one for all’ a group of people feel. Other times, operationalized in terms of how deep the friendship between members is or how well they know (and like) each other (e.g., Petersen et al., 2004), and in several organizational psychology pieces (e.g., Ayub & Jehn, 2014) defined in part as a lower likelihood of losing group members.
Social identity theory explains intergroup conflict as less of a battle for resources or power and more a striving to maintain as positively distinct, clearly defined, and optimally distinct (i.e., balanced between included and independent) the part of ourselves that is tied to important group memberships (Brewer, 1991; Brewer & Gardner, 1996; Hogg, 2012; Tajfel & Turner, 1986). Insomuch as these group memberships define us or are important, we strive to maintain positive associations with the category labels that apply to us and to do this we must have a clear idea of what defines these labels. Group membership serves many functions and identification is motivated by several drives, including comparison with others in order to learn about ourselves. In lack of an objective standard to judge our abilities and opinions, we look to groups of similar others to compare against (Festinger, 1954). Identification and subsequent comparison with similar others help people not only to get a sense of where we stand in terms of our ideas and our abilities but with validation of our self-concept as well. If comparison reflects favorably on a person, or validates our beliefs, we tend to like and identify with these groups.
People take steps to construe their self-categorized or -ascribed labels in a positive light, in other words to like who they are and be proud of their groups (Tajfel & Turner, 1986). Another motivator of group identification is self-uncertainty or specifically its reduction. There are many reasons people enter, stay in, and psychologically identify with groups, among them to maintain a positively distinct self-concept and the fact that being with similar others validates the personal ideas and traits we share with ingroup members. Self-uncertainty thus may be the prime motivator of identification, as research shows, without uncertainty, categorization alone does not as strongly predict intergroup bias (Grieve & Hogg, 1999). Before one can have a positively distinct or ‘better than’ social identity, one must know how the group is defined differently from others and similar within itself.
Identity-Uncertainty
If the self-concept or identity are the answer to the question of, ‘Who Am I?’ (Oyserman, 2001) then self- or identity-uncertainty (Hogg, 1988) is the state of being unsure about how to answer this question. Uncertainty-identity theory (Hogg, 2007) posits that identification is predicted by uncertainty related to who we are because being part of a group means there is normative or ‘popular’ support for proto- or stereo-typical traits associated with being part of a group we can reasonably self-categorize into. The major confirmed and replicated (see meta-analysis, Hogg & Choi, 2019) prediction stemming from all this is: uncertainty predicts identification, especially with self-relevant and salient groups, and more also with groups appearing to be a united entity (Hogg et al., 2007) and wherein norms or prototypes are more rigidly upheld (Hogg &Adelman, 2013). Because this identification serves ostensibly to reduce uncertainty with regards to the self-concept, the positive relationship between uncertainty and identification is stronger when the salient social category is highly entitative, or cohesive, similar, interdependent and otherwise clearly defined as a group.
In two studies, Hogg et al. (2007) found a much stronger relationship between self-uncertainty and identification in noninteractive groups described generally as similar amongst themselves and different from other groups of study participants. Uncertainty-identity theory has further been used to explain strength of identification with extremist groups and support of extremist behavior (Hogg, Adelman, & Blagg, 2010), and risky oppositional behavior (Hogg, Siegel, & Hohman, 2012). But, there has been little work (e.g., Hodson & Sorrentino, 1997; Sherman, Hogg, & Maitner, 2009) studying or manipulating these aspects of a group traits that increase self-uncertain people’s ingroup identification independently and their relationship to identification with preference combining groups where there are several norms at play that can be more or less coherent and more or less expected to correspond. Uncertainty-identity research has not conceptualized disagreement or lacking a clear modal preference as lacking traits that lend themselves to self-uncertainty reduction via identification or looked specifically at the effect of normative preference for risky choices on self-uncertain people’s willingness to interact with such groups.
Theoretically, people have a rough idea of what defines the groups they self-categorize into, and when reminded of their membership in particular categories or groups they can ascertain some idea of who they are as a member of these groups (Hogg & Choi, 2019) and how prototypical they are for these categories (Turner et al. 1987). When uncertain about who we are, we tend to show heightened identification with groups that are relevant in the moment or important to us (Mullin & Hogg, 1999), and especially when the group is clearly defined or more readily perceived as a “coherent unit” (Lickel et al., 2000, p. 224; i.e., is more entitative). For example, in ad hoc minimal groups created in an experimental setting, Hogg et al. (2007) found that participants asked to list things that made them feel unsure of how to react or behave showed heightened identification with groups perceived as more entitative because they were described as more similar amongst themselves and different from other groups. Whether the group can provide for uncertainty reduction depends generally on how much one can appropriately self-categorize into a group, and how readily it is perceived as coherently defined and a cohesive unit, as well as theoretically how clearly a prototypical position exists among members of the group.
Uncertainty can be experienced at different ‘levels’ of self: as an intra-personal need for cognitive closure or resolution of thought (Kruglanski & Webster, 1996), or as uncertainty as to what the traits defining the socially embedded self or the aspects of the self from which the existence of a social group can be inferred are. Self-categorizing leads people to take on the defining or inferred normative traits of groups they feel a part of and the less coherent the set of values or the less that such a set of values can even be inferred to exist, the less an uncertain person made cognizant of these groups would presumably identify. Identification is predicted by uncertainty in general (Hogg & Choi, 2019) but the more readily this label brings clarity of self-concept the more people tend to identify. For instance, self-uncertain people identify more with groups wherein membership is (perceived) as longer lasting (Lickel et al., 2000), where nonnormative members are more likely to be disciplined, and where the rules of membership and those governing members’ behavior are stricter and more rigidly defined (Hogg, Adelman, & Blagg, 2010). Groups that are more similar, cohesive and interdependent are examples of groups identified with more under uncertainty than their less “entitative” counterparts, these as well as groups with fewer dissenters (Marques et al., 2001) or that are otherwise more homogeneous (Peker, Hogg, & Crisp, 2010) boast more highly identified members who take more steps to improve the groups perceived standing or the clarity of the ingroup prototype.
Self-categorization theory gives the basis for the idea that self-uncertainty can be reduced by identification and outlines (see Oakes, Turner, & Haslam, 1991; Turner et al., 1987) the conditions, including normative and comparative fit, how appropriate a category is to the situation in terms of most people’s ideas about what a group is and in terms of the trait being compared between people respectively, as well as how important to a person different group memberships are, that lead to a group coming to mind in the first place. These group-level characteristics like fit with the current situation (for example divisions in athletic competitions being a valid group distinction to bring up if considering how skilled an athlete one is) or how personally relevant or chronically accessible a category is, govern whether a group will come to mind or serve to reduce feelings of self-uncertainty when they come up. Interestingly these traits are also some that lead to perceptions of a group as entitative, for example if observing frequent interaction between group members it would be less of a surprise that they belong to the same social category, and the frequency with which a person interacts with a group would also make it more likely to come to mind as a source of uncertainty reduction. If the definition of the category is clear or presumed to be so, due respectively to lack of dissent or the presence of homogeneity, one of the traits that supplant perceptions of a group as a single entity (Lickel et al., 2000), identification with this group would reduce feelings of subjective self-uncertainty more effectively. Working in a group would make people in the work group salient as a social category in the moment and a candidate for uncertainty reduction through identification.
This concept of perceived entitativity is multifaceted and most easily defined as the extent to which a collection of things is seen as one entity (Campbell, 1958). Perceptions of entitativity are supported by things like how impermeable or easy to join a group is, how long it has lasted or been around, how cohesive people are, how similar they are, how interdependent their actions, and how intertwined their fates (Hamilton & Sherman, 1996; Lickel et al., 2000). There are several dimensions of entitativity, several qualities that lead to the perception of a group as one thing, and these different aspects are not necessarily correlated themselves. For instance, Lickel et al. (2000) found that while importance and permeability ratings of a wide array of types of groups were moderately correlated (r = -.37), between the two, importance is more strongly associated (r = .41) with perceived entitativity than is permeability (r = -.24). Traits making a group an ideal candidate for identification when self-uncertain need not always be related. The actual content of a group’s norms or what marks an ingroup member as prototypical can be more, or less, difficult to find or define, and more-or-less related to other traits groups have in common. If deference to an ingroup prototype is the mechanism for identification reducing uncertainty, then lack of a prototypical position might also be less than ideal for self-uncertain members of this group, unless the group is perceived as entitative and able to find consensus from disagreement based on other traits.
A Note on “Diversity”
Intra-group differences in task-oriented groups that may have different ideas for how to complete a task are also studied in the context of organizational work groups (Hogg & Terry, 2000) and “diversity.” Diversity is heterogeneity on any trait that can be used to differentiate between people in a group (Jackson, 1992; Triandis et al., 1994; Williams & O’Reilly, 1998). Harrison and Klein (2007) differentiate between three types of diversity: variety diversity, referring to intra-group differences on one trait, separation diversity, which is any heterogeneity that can act as a cue to placing people into qualitatively different subgroupings, and disparity diversity, referring to inequality between group members. Others (e.g., Phillips, Northcraft, & Neale, 2006) also distinguish between surface-level or superficially apparent traits like height or broadcasted traits, which can include traits that otherwise would be deep-level traits that are typically not self-evident, like differences in personal values. By now, the proposition that any and all types of diversity can have negative or positive effects on group cohesion and performance (van Knippenberg, Homan, & De Dreu, 2004) has gained much traction. Because different dimensions of diversity can be more-or-less correlated in different contexts; it is not the type of divides present but whether or not diversity highlights intergroup distinctions or fault lines (Lau & Murnighan, 1998) across which different task-relevant information or opinions are distributed, that decreases identification with and performance quality of groups on tasks that necessitate the broadest survey of information or working through a difference of opinions (Homan et al., 2007). Diversity defined broadly as heterogeneity on any intra-group trait can alternatively be operationalized as the presence in a group of dissent, nonnormative members, disagreement, a minority voice, or anything that generally undermines or makes consensus unclear.
The subjective group dynamics model explains reactions to members holding non-prototypical opinions and negative evaluations of such members as normative members trying to reaffirm the validity of subjective norms by devaluing and sending communications to that effect to members who endorse nonnormative opinions and bring this validity into question by breaking consensus (Eiedelman, Silvia, & Biernat, 2006; Frings, Abrams, Randsley de Moura, & Marques, 2010; Marques, Abrams, & Serodio, 2001). The ingroup projection model also suggests diversity is a threat to self-definition, as the tendency to project subgroup traits onto the superordinate category presumably in order to paint the in-subgroup as more prototypical and worthy of unequally distributed resources (Waldzus et al., 2003; Wenzel, Mummendey, & Waldus 2009) is significantly decreased as is identification with the superordinate category when it is described as complex or having multiple prototypes (Peker, Crisp, & Hogg, 2010). Complex social categories described as having multiple discernible norms or no single discernible norm or group definition, tend to be identified with less. These are both examples of disidentification and disliking of the ingroup motivated by there being diversity defined broadly as a lack of a clear prototype or group definition.
Several lines of evidence demonstrate the highest levels of identification are exhibited by members whose complimentary distinctions within a shared superordinate category and their shared category membership are simultaneously emphasized (Deschamps & Brown, 1983; Hornsey & Hogg, 2000; van Leeuwen, van Knippenberg, & Ellemers, 2003), and that threats to positive distinctiveness, and theoretically then, self-uncertainty made salient, make these tendencies more pronounced (Brewer & Gardner, 1996). For example, when complimentary roles are given to task group members and their performance collectively evaluated, making salient superordinate category membership, identification is higher on average than when roles are not assigned and group goals are not emphasized (Brown & Wade, 1987). A group of people believing something or acting a certain way is often assumed to provide some subjective evidence for these beliefs or actions being correct, and this also explains some behavior in decision-making groups. A person being unsure about who they are might remind themselves or a get a clearer answer to that question by thinking of the ideas collectively held by groups that they are assigned to or feel a part of, especially if they are similar or potentially in agreement.
Task oriented group members provide a special case where all members share a common goal, and may be more or less on the same page as far as the specifics of what the group is supposed to accomplish, but may vary on how homogeneous they are on traits related to different degrees to (preference for) the options a group can choose from and in terms of the pre-interaction preferences these individuals have for how or what to choose. Norms are what are perceived and observed as the average tendencies of groups (Cialdini, Reno, & Kallgren, 1990) and prototypes an accentuation of these norms. Social norms are averages but as such do not represent the exemplary group member’s position and do not by definition represent extreme positions. This distinction matters because some behaviors characteristic of certain groups can also ostensibly reduce feelings of self-uncertainty, like those that show extreme loyalty or dogmatic adherence to identity-central beliefs, serving potentially to uphold the group as entitative, as can simply being a member of a group easily perceived as entitative.
Uncertainty & “Diversity” in Work Groups
Being in a task group may be enough to emphasize group boundaries and lead to inter-group bias and ingroup identification (Asch, 1956; Sherif, 1958), especially in combination with uncertainty (Hogg & Grieve, 1999), but what happens when the task at hand is specifically to find consensus and the group is diverse and or clearly has some ways to go before agreeing on anything? Is uncertainty like a threat to positive distinctiveness in that simultaneous emphasis on intra-group similarity is needed to balance out differences of opinions among self-uncertain task group members? If so, will simply being in a potentially interactive group described as dissimilar count as simultaneously emphasizing superordinate category membership and highlighting complimentary distinction, or is some sort of obvious homogeneity or potential for agreement needed? Identification might be highest with agreeing or similar groups, not agreeing and similar, because of the need to balance potential subgroup distinctions and group membership as per evidence cited above. However, if identification is serving to reduce self-uncertainty, similarity most likely will not be perceived as a to positive distinctiveness if there are no differences on important self-evaluative dimensions or traits, as is the case with task-irrelevant traits in groups discussing subjective and low-stakes opinions, then it is more likely that both agreement and homogeneity will bolster identification.
The norm in a decision-making group or a group tasked with coming to a single choice or of individuals working in such a task-oriented consensus-seeking group is the average of the individuals’ preferences or when dealing with categorical response scales or option sets the majority-preferred response (the mean and mode being more similar the less a distribution of preferences is skewed). Distinguishing between the average tendency and the prototypical tendency, which is an exaggeration of the average and typical traits of members of a specific group, is important; leaders are typically perceived as what some (e.g., Marquez et al., 1988) call pro-norm deviants because they follow group norms more than the average person and are looked to as upholding the definition of a group (Abrams et al., 2008; Giessner et al., 2009). Leaders or prototypical members are not just people who represent the average tendency of the group but are the best representatives and are more influential to highly identified ingroup members (Haslam, Reicher, & Platow, 2011). Similarly, people who are the most central members of their group would be the biggest proponents of beliefs that define a group, for instance a priest is not expected to just sort of believe in his religion. If we look at a response distribution of answers to a religiosity survey, to find the priest’s score, you would not look at the average, you would check to see if his was one of the highest religiosity scores. By the same token, a group more clearly believes something if their average is skewed, especially if being in the middle means being unsure and not just lukewarm about something.
There may be a direct link between demographic diversity and clarity of ingroup prototypical positions even in artificially created work or task-oriented groups. Chattopadhyay, George, and Lawrence (2004) for instance demonstrated that the proportion of minority (female or non-citizen) members in a work group negatively correlated not only with positive evaluation of the group, as well as perceived clarity of group definition, but also with perceptions of the self as prototypical. That is, diversity predicted lower group esteem (i.e., the reason one would identify was lower), more uncertainty with regards to group definition, and lowered perceptions of individual work group members’ status within the group, perhaps further implying that heterogeneity muddling group definition also leaves individual members without a means for taking on prototypical traits themselves and this affects how positive it is to be a group member. Decision-making in groups is the process of individuals choosing from two or more response options in contexts where there is not necessarily but can be an objectively or demonstrably correct answer (Kerr & Tindale, 2004). Small group decision-making or work in the context of preference combining groups is essentially the process by which individual group members’ possibly diverse or diametrically opposed preferences about which of two or more options is best and the array of arguments and expertise of different people are optimally combined into a consensually agreed upon and most-informed collective choice. Hence, cognitive processes related to combining preferences, skills, and information as well as social processes related to intra- and inter-subgroup differentiation, comparison, and cooperation (e.g., cohesion, perceived potential relational, task, or process conflict, and their interaction or lack thereof), play significant roles not only in performance but satisfaction with the group, as well as group member identification, and potential group turnover.
Self-uncertainty specifically being theoretically and empirically related to people’s search for and stricter adherence to clear norms or prototypes and their acceptance or aversion to heterogeneity should play a role in identification with preference combining groups. Interestingly, even in groups like juries instructed to come to a unanimous decision, the collective single choice can be predicted from the decision that most people already preferred individually (Kerr & Tindale, 2004). Consider the following scenario, a perfectly hung jury with six in favor of a conviction, and six against: Self-uncertain jurors at least have their identity as jurors with a clear task at hand to think about as a means for reducing their uncertainty (though perhaps not so much uncertainty related to jurors’ self-concepts rather than the task at hand) and from there, anything that makes the group appear more a single entity with any number of traits on any dimension that lead to perceptions of groups as entitative would further reduce uncertainty. In lack of any of these traits, a hung jury is one that (a) does not have a clear path to consensus, and (b) is specifically lacking in a preference norm as the average between ‘yay’ and ‘nay’ or conviction and reasonable doubt is ambivalence or the ‘jury’s still out’, so also (c) if they share no discernible traits of any kind, identifying with the role juror alone might not bring to mind much of a clear picture as to what defines a person in this context, relative to a jury that either is leaning one way or the other (i.e., sharing a polarized normative or prototypical position) or shares even verdict-irrelevant traits (e.g., all employed). Furthermore, without anything to make the group seem more like a unit, uncertainty-identity theory and SCT both imply that not just any norm will do to reduce uncertainty and a skewed norm or prototype (and further one skewed in favor of more punitive ideas) not just an average to adhere to is necessary for the relative reduction of self-uncertainty.
It is conceivable that a skewed norm is needed if similarity or other traits lending themselves to the perception of a group as entitative are not present, for identification to reduce uncertainty; similarity combined with obviously being in a group of twelve might do it (though the question of the present research is whether, given a minimal group scenario, similarity or a clear preference norm or the combination of the two versus the lack of either will lead to higher relative identification and attraction to decision-making groups, all people being in a task group). In this example, the implications of this research program are highlighted: groups like juries are likely to at least be task-uncertain and anything that makes them work together better under these circumstances is obviously ideal as is knowing what leads people to avoid more complex debate. The norms here that uncertain people may be more prone to adhere to are (1) the larger norms dictating a lack of chair throwing and avoiding a ‘12 Angry Men’ or entrenched debate situation, (2) any similarity that can be but does not have to be related to the average preference of a group, and (3) the prototypical position or the average position if there is no a priori expected direction of the preference of a group (Turner et al., 1990). These three all interact the more they overlap so that if a preference combining group is a group of people also having another trait in common, similar in a way that suggests their average preference on a scenario or attitude, the more under uncertainty people would theoretically go along with and identify more with a similar and agreed decision-making group.
Whether the tendency in this and other lines of research (e.g., groupthink, shared information bias) to lean into the predicted or perceived attitudinal or preference norm of a decision-making group increases in self-uncertain people viewing a group potentially seen as low in entitativity (i.e., dissimilar), as it theoretically does in the presence of positive distinctiveness threats (van Knippenberg et al., 2004) has not been studied. Group polarization, for example, is the tendency of discussion groups to respond together more extremely than their average responses prior to discussion. It rests both on the existence of more arguments for the average position (persuasive arguments theory, Burnstein & Vinokur, 1977) and the validity of these arguments being trusted because they are supported by the majority of potential discussion group members (normative social influence model, Myers et al., 1981). The former suggests there are more arguments in favor of the normative position that could be brought up in conversation, convincing more people to follow the norm, the latter that simply knowing the majority of a group tends a certain way means the average position is probably true or at least is the way to go to fit in. The self-categorization take on the phenomenon has it that these two factors are inextricably linked, that when something is validated by consensus, there are more likely more arguments presumed to exist and that the average response of a group is not one that leans one way or another necessarily without a social frame of reference.
Point being, polarization research presents many motives for trusting the veracity of or otherwise going along with opinions that are presumed to be validated by a majority given an average response that is different than the midpoint of a response scale and a social context that presumes agreement within a group and disagreement between groups. If self-uncertainty leads to a search so to speak for cues to there being clarity of group mind, purpose, or definition and identification with such groups, there should be more identification with groups with polarized norms presumed to be more valid. Polarization research became popular after a seminal dissertation study and follow-up publication (Stoner, 1961; 1968) giving evidence of a supposed risky shift that gave way to describing the general phenomenon of discussion groups accentuating norms, as the norm before discussion that is figured out during discussion is not always leaning risky because the population of the planet does not lean in favor of risk generally. Though there was at first, now there is no presumed societal norm of risk-taking; some groups may be more risk-averse than others, but the average experimental group is no longer expected to lean toward risky choices. That said, risk as operationalized in these papers includes some of the behaviors described as furnishing a more rigid sense of what a group believes as per a theory piece arguing for the uncertainty-reducing properties of delinquent behavior (Hogg, Siegel, & Hohman, 2012). in addition to this work and the finding that uncertainty predicts higher endorsement of inter-group violence (Hogg et al., 2010), there is also the fact that ‘caution’ has a connotation of playing it safe if unsure whether risk will pay off, yet there has not been an experimental test of whether viewing a group preferring risk-taking would reduce uncertainty via identification more effectively than a group mostly leaning toward a cautious option.
Rarely, if ever, in naturalistic settings does a group’s average fall right on the midpoint of a response scale, especially if the middle option is a neutral response; just as rarely does a predetermined number of coin flips come out exactly half heads-half tails so neither came up more frequently. Outside of probabilities, because people do not live in a vacuum and there is usually a relative frame of (social) reference at least implicitly present, choice shifts are less a skewing of an un-skewed distribution of individual preferences and more adherence to a norm polarized away from relevant outgroups and the more one identifies for any reason the more groups will show they prefer a position shifted away from the middle of a normal curve. The average tendency of a group is not necessarily the same as a prototypical, ideal, or exemplary tendency and being an average member is not being the epitome of a group member so the need to reduce uncertainty should lead to a preference for skewed rather than evenly distributed preferences in the group. However, it is possible more so with people that stalemates of a sort occur where there is obviously an average but it is at the midpoint between two options or between extreme endorsement and slight endorsement of an attitude or choice, implying that each response option is preferred by the same number of people and then it purely boils down to other factors of who has more convincing arguments, who is more persuasive, and who is more motivated to have their response win out. Point being, there is always an average response or norm but not always a clear collectively preferred response in groups that must combine preferences. An average or mean preference can be in between two possible courses of action and because an equal number and together the majority of people prefer one or the other. If the mechanism for uncertainty reduction is being in a group that can be presumed to have or has a polarized norm, then self-uncertain people should identify more with non-ambivalent groups whose average preference is skewed toward one course of action or another especially when a group is otherwise diverse.
Summary
Self-uncertainty leads to higher identification with homogeneous and easily defined groups as a means for reducing this uncertainty and people in decision-making groups, which can be homogeneous in terms of preferred decision and group member characteristic, tend to accentuate modal preferences especially in light of other factors increasing (perceived) group cohesion (e.g., homogeneity). Identification with groups whose preferences or attitudes are central to their purpose also increases the tendency to uphold (Marques et al., 2001) and run with a perceived norm or agree with a normative preference (e.g., Cialdini & Trost, 2001; Livingstone, Haslam, Postmes & Jetten, 2011) presumably because these norms are validated or made more certain by consensus. The question is if uncertainty then would lead people to identify more with a group leaning one way or another in terms of task-relevant attitudes or choice preferences especially in groups tasked with reaching consensus, especially when there is no other aspect of the group that helps particularly to reduce uncertainty. Those factors such as threats to the positivity and uniqueness of relevant self-ascribed category labels and feelings of self-uncertainty that affect identification or group relevant behavior (e.g., positive distinctiveness threats leading to bias and uncertainty to disidentification with low entitative groups) respectively do and should correlate with the observation of bias in the direction of the norm of preference-combining groups, and tendencies to disidentify with and psychologically ostracize nonnormative ingroup members that make a prototype or polarized norm unclear. Several phenomena, groupthink, the tendency to focus on the common information or the favored response in a consensus-seeking group, all have a social identity-relevant component to them and therefore self-uncertainty is a logical candidate as an important predictor of identification with groups falling prey to these tendencies.
Identification predicts groupthink like tendencies (Turner & Pratkanis, 1994; Turner et al., 1992) and Hodson and Sorrentino (1997) found evidence that groupthink or dogged pursuit of consensus to the detriment of group performance was more likely to occur among people with a lower tolerance for uncertainty. Van Knippenberg and colleagues’ (2004; 2007; 2008; 2011) work on the categorization-elaboration model suggests that heterogeneity combined with variables that increase the need for identification with homogeneous groups like threats to positive distinctiveness (therefore also how well-defined a group is) dovetail with the tendency to avoid discussing unshared and potentially conflicted information. Lastly, the social identity take on interaction of diverse subgroup members (see Hornsey & Hogg, 2000 for a review) and reactions to nonnormative group members (Marques, Yzerbyt, & Leyens, 1988) as well as group polarization research (Hogg, Turner, & Davidson, 1997) suggest the need for simultaneous emphasis on complimentary intra-group differences and collective goals for group members to remain identified, and that accentuation of preference norms also relates to being able to ascertain these norms in order to follow them. All lines point to the fact that differences of opinions will not be so well tolerated by self-uncertain people unless there are other things about the consensus-seeking group to balance out the lack of a clear direction and that disagreement will not be as threatening to people seeking consensus if there is intra-group similarity to at least make the group appear more likely to reach consensus.
People are persuaded more by group-prototypical positions and like arguments for prototypical attitudes more (van Knippenberg & Wilke, 1992) because they are seen as more potentially valid and would therefore reduce subjective uncertainty about them and increase identification in uncertain people with a group they are about to discuss their position with. However, if uncertainty is reduced by perceptions of the group as similar within itself, much in the same way bias flowing from self-categorization processes is mitigated and leads to more sharing of disparate information and opinions, in a collective choice context, people may be more likely to at least consider differing opinions, like reducing the need to fit in generally does lead to more willingness to share potentially wrong answers (Elwyn et al., 2012; Politi & Street, 2011). The present research assesses whether uncertainty will lead people to be more cohesive and identified with groups with an obvious majority preference, especially when the group is described as dissimilar personality-wise, while possibly identifying more and being more willing to work with people who are similar but have no obvious majority preference pre-discussion.
Alex Ash Creative 