Baumeister et al. devote much space to reviewing well-established research that purportedly illustrate paradoxical findings related to group formation and group performance, but fall short of moving the field forward. Much discussion centers on accepted wisdom that group members emphasise cohesiveness early in their formation, that deindividuation reduces accountability and usually produces unsavoury outcomes, as well as other maladies of group processes. More interesting, Baumeister et al. argue that such phenomena illustrate the value of differentiated selves. However, we found this focus on differentiated selves uninformative in providing new insights about the apparent paradoxical relationships between social facilitation and social loafing in a way that reaches beyond decades-old research on the importance of identifiability of group members' contributions (e.g., Harkin Reference Harkins1987; Karau & Williams Reference Karau and Williams1993).
As McGuire (Reference McGuire2013) observed, a key goal of research is to identify conditions that give rise to specific outcomes. Despite briefly acknowledging situations in which differentiated selves may hinder group performance, the target article is mostly a panegyric of such differentiation, which we see as a very selective read of extensive literature. This may be captured most clearly by what we consider the crux of the target article's framework, the sweeping claim that “the worst outcomes of group processes come when individual identities are submerged in the group. By submerged in the group, we mean any of the following: People are held neither accountable nor responsible, they are not in competition nor playing a distinct role, they are not publicly identified nor rewarded” (para. 5, emphasis on any ours). We propose that this one-sided account is an outcome of two shortcomings. First, in using the construct differentiated selves, the authors collapsed identifiability and specialization – which we believe are two distinct elements. Second, the framework is not refined enough to provide an appropriate foundation for understanding group-related outcomes.
The first main weakness is the creation of a single construct, differentiated selves, from two distinct ones. The quote above illustrates this by capturing under a single umbrella identifiability (“publicly identified”) and specialization (“playing a distinct role”). We argue that not only are these elements conceptually different, but also there are conditions under which group outcomes will be improved when either is high or low. For example, as the target article reviewed at length, identifiable contributions increase accountability, allowing for improved outcomes at times; however, as research indicates, one of the symptoms of the undesirable groupthink is pressure on dissenters to acquiesce, which is effective mostly when the identity of the dissenters is known (Janis & Mann Reference Janis and Mann1977). Similarly, while specialization often offers distinct advantages to group productivity (e.g., Stasser et al. Reference Stasser, Stewart and Wittenbaum1995), it also places a lot of power in the hands of specific individuals with critical expertise, leaving the group vulnerable to potential demands of such individuals or when they depart the group.
Whereas the first weakness was a term-specific one, the second weakness relates to the crux of the framework as summarised in the quote above. We find this framework to be an oversimplification, inadequately reflecting thoughtful previous research in many areas, while potentially stifling future research as such a framework does not easily lend itself to the creation of refined novel hypotheses. Instead, we argue that adopting a framework of a group-by-situation (GxS) interactive perspective enables us to advance our understanding of group processes through the focus on group related variables in the context of changing landscapes, allowing for more accurate predictions to transpire.
To substantiate this critique we can consider how the GxS promotes innovative research whereas the target article's framework may stifle it. The proposition that after a stage of group building (cohesiveness) one would benefit from increased differentiation (of both identifiability and specialization presumably) neglects many common findings and moreover may fail to allow such research to flourish. To illustrate, consider research on group conflict and performance. Early models suggested that conflicts interfere with effective group performance (e.g., Hackman & Morris Reference Hackman, Morris and Berkowitz1975; Pondy Reference Pondy1967). These approaches were later criticized for failing to account for the benefits of task (as opposed to relational) conflicts, which are “conflicts about the distribution of resources, procedures and policies, and judgements and interpretations of the facts” (De Dreu & Weingart Reference De Dreu and Weingart2003, p. 741). Following the target article's crux, it is likely that task conflicts will arise with increased identifiability as both are considered to encourage greater cognitive understanding and investment in the issues considered. However, despite the potential attractiveness of the underlying argument, a meta-analysis of the relevant research indicates that “for team performance, both task conflict and relationship conflict are equally disruptive” (De Dreu & Weingart, p. 746). Such findings not only undermine the differentiated selves' framework, but also they illustrate the target article's failure to account for many nonsupportive findings.
Just as the research on team conflict undermines the suggestion that identifiability will (almost) always improve group outcomes, other research undermines the claim that specialization will always contribute to performance. For example, a recent meta-analysis on shared leadership (which reflects decreased specialization, as an important role is occupied by multiple individuals) demonstrates its overall positive association with team performance (D'Innocenzo et al. Reference D'Innocenzo, Mathieu and Kukenberger2014).
Limited space does not allow us to elaborate on the GxS perspective that we favor. In short, we contend that identifying relevant group variables – such as goals (e.g., ad-hoc vs. continuing; achievement vs. ideological; maintenance vs. innovation), structure (hierarchical vs. egalitarian; small vs. large; open vs. closed), dynamics (e.g., team climate; West & Anderson Reference West and Anderson1996, affective tone; George Reference George1990), and traits (e.g., group-level traits; Smaldino Reference Smaldino2014, members' traits; LePine et al. Reference LePine, Hollenbeck, Ilgen and Hedlund1997) – in a context of specific situations (e.g., fast changing vs. stable; recurrent vs. novel; relational vs. goal-focused; simple vs. complex) will promote a more accurate and generative foundation for the identification of optimizing team outcomes. We believe that nascent research in this direction already exists (e.g., De Dreu et al. Reference De Dreu, Nijstad and van Knippenberg2008) and would recommend further developing these models to capture the complexities that are absent from the framework offered in the target article.
You have
Access
Baumeister et al. devote much space to reviewing well-established research that purportedly illustrate paradoxical findings related to group formation and group performance, but fall short of moving the field forward. Much discussion centers on accepted wisdom that group members emphasise cohesiveness early in their formation, that deindividuation reduces accountability and usually produces unsavoury outcomes, as well as other maladies of group processes. More interesting, Baumeister et al. argue that such phenomena illustrate the value of differentiated selves. However, we found this focus on differentiated selves uninformative in providing new insights about the apparent paradoxical relationships between social facilitation and social loafing in a way that reaches beyond decades-old research on the importance of identifiability of group members' contributions (e.g., Harkin Reference Harkins1987; Karau & Williams Reference Karau and Williams1993).
As McGuire (Reference McGuire2013) observed, a key goal of research is to identify conditions that give rise to specific outcomes. Despite briefly acknowledging situations in which differentiated selves may hinder group performance, the target article is mostly a panegyric of such differentiation, which we see as a very selective read of extensive literature. This may be captured most clearly by what we consider the crux of the target article's framework, the sweeping claim that “the worst outcomes of group processes come when individual identities are submerged in the group. By submerged in the group, we mean any of the following: People are held neither accountable nor responsible, they are not in competition nor playing a distinct role, they are not publicly identified nor rewarded” (para. 5, emphasis on any ours). We propose that this one-sided account is an outcome of two shortcomings. First, in using the construct differentiated selves, the authors collapsed identifiability and specialization – which we believe are two distinct elements. Second, the framework is not refined enough to provide an appropriate foundation for understanding group-related outcomes.
The first main weakness is the creation of a single construct, differentiated selves, from two distinct ones. The quote above illustrates this by capturing under a single umbrella identifiability (“publicly identified”) and specialization (“playing a distinct role”). We argue that not only are these elements conceptually different, but also there are conditions under which group outcomes will be improved when either is high or low. For example, as the target article reviewed at length, identifiable contributions increase accountability, allowing for improved outcomes at times; however, as research indicates, one of the symptoms of the undesirable groupthink is pressure on dissenters to acquiesce, which is effective mostly when the identity of the dissenters is known (Janis & Mann Reference Janis and Mann1977). Similarly, while specialization often offers distinct advantages to group productivity (e.g., Stasser et al. Reference Stasser, Stewart and Wittenbaum1995), it also places a lot of power in the hands of specific individuals with critical expertise, leaving the group vulnerable to potential demands of such individuals or when they depart the group.
Whereas the first weakness was a term-specific one, the second weakness relates to the crux of the framework as summarised in the quote above. We find this framework to be an oversimplification, inadequately reflecting thoughtful previous research in many areas, while potentially stifling future research as such a framework does not easily lend itself to the creation of refined novel hypotheses. Instead, we argue that adopting a framework of a group-by-situation (GxS) interactive perspective enables us to advance our understanding of group processes through the focus on group related variables in the context of changing landscapes, allowing for more accurate predictions to transpire.
To substantiate this critique we can consider how the GxS promotes innovative research whereas the target article's framework may stifle it. The proposition that after a stage of group building (cohesiveness) one would benefit from increased differentiation (of both identifiability and specialization presumably) neglects many common findings and moreover may fail to allow such research to flourish. To illustrate, consider research on group conflict and performance. Early models suggested that conflicts interfere with effective group performance (e.g., Hackman & Morris Reference Hackman, Morris and Berkowitz1975; Pondy Reference Pondy1967). These approaches were later criticized for failing to account for the benefits of task (as opposed to relational) conflicts, which are “conflicts about the distribution of resources, procedures and policies, and judgements and interpretations of the facts” (De Dreu & Weingart Reference De Dreu and Weingart2003, p. 741). Following the target article's crux, it is likely that task conflicts will arise with increased identifiability as both are considered to encourage greater cognitive understanding and investment in the issues considered. However, despite the potential attractiveness of the underlying argument, a meta-analysis of the relevant research indicates that “for team performance, both task conflict and relationship conflict are equally disruptive” (De Dreu & Weingart, p. 746). Such findings not only undermine the differentiated selves' framework, but also they illustrate the target article's failure to account for many nonsupportive findings.
Just as the research on team conflict undermines the suggestion that identifiability will (almost) always improve group outcomes, other research undermines the claim that specialization will always contribute to performance. For example, a recent meta-analysis on shared leadership (which reflects decreased specialization, as an important role is occupied by multiple individuals) demonstrates its overall positive association with team performance (D'Innocenzo et al. Reference D'Innocenzo, Mathieu and Kukenberger2014).
Limited space does not allow us to elaborate on the GxS perspective that we favor. In short, we contend that identifying relevant group variables – such as goals (e.g., ad-hoc vs. continuing; achievement vs. ideological; maintenance vs. innovation), structure (hierarchical vs. egalitarian; small vs. large; open vs. closed), dynamics (e.g., team climate; West & Anderson Reference West and Anderson1996, affective tone; George Reference George1990), and traits (e.g., group-level traits; Smaldino Reference Smaldino2014, members' traits; LePine et al. Reference LePine, Hollenbeck, Ilgen and Hedlund1997) – in a context of specific situations (e.g., fast changing vs. stable; recurrent vs. novel; relational vs. goal-focused; simple vs. complex) will promote a more accurate and generative foundation for the identification of optimizing team outcomes. We believe that nascent research in this direction already exists (e.g., De Dreu et al. Reference De Dreu, Nijstad and van Knippenberg2008) and would recommend further developing these models to capture the complexities that are absent from the framework offered in the target article.