L EADER P ROTOTYPICALITY AND L EADER A CCEPTANCE : T HE ROLE OF PREVIOUS GROUP PERFORMANCE
Master’s Thesis
MSc Human Resource Management University of Groningen Faculty of Economics and Business
January 16th, 2016
Eugenia Fink Student number: S2826364
Gratamastraat 34 Tel.: +31 (0)633-602856
9714HN Groningen E-Mail: e.fink@student.rug.nl
Supervisor: Prof. Dr. B.A. Nijstad
TABLE OF CONTENTS
INTRODUCTION ... 1
LITERATURE REVIEW ... 3
Group Prototypicality ... 3
Previous Group Performance ... 7
METHOD ... 10
Participants and Design ... 10
Procedure and Manipulations ... 10
Measures ... 13
RESULTS ... 16
Manipulation Check ... 16
Hypotheses Testing ... 18
Additional Analysis ... 20
DISCUSSION ... 23
Implications ... 27
Limitations and Future Research ... 28
CONCLUSION ... 31
REFERENCES ... 32
LIST OF FIGURES
FIGURE 1 – Conceptual Model ... 9
LIST OF TABLES
TABLE 1 – Experimental Conditions ... 10 TABLE 2 – Manipulation Check ... 16 TABLE 3 – Descriptive Statistics ... 17 TABLE 4 – Effects of Moderated Mediation Analysis (Independent Variable: Manipulated Prototypicality) ... 19 TABLE 5 – Effects of Moderated Mediation Analysis (Independent Variable: Perceived Prototypicality) ... 21
ABSTRACT
The group prototype represents beliefs of group members about characteristics of the ideal group member. Accordingly, the more prototypical a person is, the more he or she fulfills the image of this ideal member. Prototypical group leaders are typically more positively evaluated by their group members than non-prototypical group leaders. How this phenomenon is affected by previous group performance in case of leader change was evaluated in the present study. A small group experiment was conducted to test the hypotheses in which we manipulated leader prototypicality and task performance of the group. Findings showed positive effects of prototypicality on perceived competence and acceptance of the leader.
Effects of prior group performance showed the tendency that prototypical group leaders are even more positively evaluated than non-prototypical group leaders after a leadership change.
Expectation States Theory and Social Identity Theory are used to elaborate on the findings.
INTRODUCTION
The effectiveness of group leaders is of crucial importance to various group outcomes such as group productivity, satisfaction (Kahai, Sosik, & Avolio, 1997), creativity (Kahai, Sosik, & Avolio, 2003), and performance (Sicasubramaniam, Murry, Aviolio, & Jung, 2002).
Whether group leaders have the capability of leading the team effectively depends on characteristics of the leaders and the extent to which the group members view them as competent and accept their leadership. A matching process between the leader’s characteristics and the group members’ needs and characteristics is of significant importance for the optimal group performance and success (Pierro, Cicero, van Knippenberg, &
Kruglanski, 2005). The Social Identity Theory of Leadership aims at explaining which group processes influence how leaders emerge in a group and how group members react to them.
According to this theory, this depends on group prototypicality. Prototypes are regarded as characteristics that distinguish one social group from another based on beliefs and perceptions of the group members. The prototype changes accordingly, depending on the reference group, as groups can hold different beliefs (Hogg & van Knippenberg, 2003). The more prototypical an individual is, the more he or she represents the ideal group member (Hogg & Terry, 2000).
One reason why prototypical group leaders are more accepted is that the group members expect someone who is similar to them to act more in their favor than someone who differs from the group (Hogg & van Knippenberg, 2003; van Knippenberg & van Knippenberg, 2005). However, leaders should also be perceived as competent to be accepted as group leaders. For example, Expectation States Theory emphasizes the importance of expectations about leader competence (Correll & Ridgeway, 2006; Rink & Ellemers, 2008).
As prototypical group members are often regarded as being more competent, it could be assumed that prototypical group leaders are always preferred. However, we suggest that this relationship depends on the previous performance of the group. In case of a leadership change
previous group performance needs to be considered when identifying necessary characteristics of the subsequent new group leader. The arising research question is the following: How does previous group performance influence the relationship between leader- group prototypicality and perceived competence of the leader and how does this relate to the leadership acceptance?
Previous findings suggest that prototypical group leaders are more successful and more accepted than non-prototypical group leaders (Cicero, Pierro, & van Knippenberg, 2009; Hogg & van Knippenberg, 2003; Pierro et al., 2005; van Knippenberg, 2011). However, if the group performed poorly on its previous tasks with a prototypical leader, we propose that it is likely that they will value a non-prototypical leader more in case of a leadership change.
The self-image of the group is likely to be more negative after a poor performance. Therefore, it is likely that a prototypical new leader, who would be regarded as part of the in-group, also carries this negative image. A non-prototypical leader would therefore seem desirable as he or she does not inherit the negative stigmatization.
The extent to which leader prototypicality affects the leader’s perceived competence and how, in turn, that influences leadership acceptance will be evaluated in this study.
Furthermore, prior performance of the group will be taken into account. It will be examined how previous group performance influences the relationship between perceived group prototypicality and the leader’s perceived competence in case of leadership-change. These relationships will be evaluated in an experiment manipulating prior performance of the group by measuring the impact of leader prototypicality and previous group performance on perceived leader competence and leadership acceptance using standardized questionnaires.
LITERATURE REVIEW
Group Prototypicality
Group leaders function as part of the social group in which they work. Thus, it is crucial to not only think of them as a separate entity but also in terms of group dynamics. In social groups, all group members are evaluated by one another in terms of their degree of prototypicality with regard to the group. The more an individual shares perceptions and beliefs of the respective group, the more he or she is regarded as being prototypical (van Knippenberg, 2011). This evaluation may change when the reference group changes, as cognitive features are group specific and may develop over time. Accordingly, new individuals entering a social group may alter the existing group prototype (van Knippenberg
& van Knippenberg, 2005). With that, also the person who is perceived to be the most prototypical group member can vary. The extent to which the prototype changes is dependent on the size of the group and the type and amount of group membership change (Hogg, 2001;
Hogg & van Knippenberg, 2003).
The group prototype does not represent the average group member but the perceived ideal group member, according to a specific social group. Respectively, the more prototypical an individual is the more he or she fulfills the expectations of his or her group members (van Knippenberg & van Knippenberg, 2005). It is consequently not surprising that prototypical group members often emerge as leaders of a group and that prototypical group leaders are more accepted and regarded as being more effective, as they represent the group ideal to a greater extent (Pierro et al., 2005; van Knippenberg, 2011).
The extent to which the group leader is perceived to be prototypical affects the acceptance of the leadership position because prototypical group members are generally expected to act more in favor of their social group, as they are expected to ideally share morals and beliefs. On the contrary, non-prototypical group leaders do not have this initial
advantage and need to prove to the group that they are capable of acting in favor and in the interest of the group (van Knippenberg & van Knippenberg, 2005). Furthermore, prototypical group leaders are perceived to act more just than non-prototypical group leaders which is why actions of non-prototypical group leaders are judged more strictly than actions of prototypical leaders (Platow & van Knippenberg, 2001).
Tanghe, Wisse, and van der Flier (2010) have shown that identification with a group enhances the willingness of individuals to help their group members and also the extent to which “group members are open to the influence of other group members” (Tanghe et al., 2010: 353). As identification with the group is related to group prototypicality, team performance was found to be higher if group prototypicality is salient. More prototypical group leaders were found not only to be more accepted by their group members but also to be more effective (van Knippenberg & Hogg, 2003). Furthermore, as the most prototypical group member represents the most ideal member in the group, other group members may strive to reach this ideal. It therefore seems that leaders, as they are perceived to be the most prototypical, exert influence over the other group members because these strive to conform to their model (Hogg & Terry, 2000).
Group members that identify with the group are more willing to pay attention to the verbal and non-verbal communication of other group members. The possibility to influence group members augments when the degree of prototypicality is higher, as the group members are more attentive to the signals of their prototypical group leader (Tanghe et al., 2010).
Accordingly, prototypical group leaders are perceived to inherit more influence over group members than non-prototypical leaders (Hogg, 2001). It is therefore possible that prototypical group leaders are merely perceived as being more effective and competent because the other group members are more open to their influence.
Considering that the performance evaluation of prototypical leaders is more positive and that the possible influence over team members also increases with group prototypicality, our first hypothesis is as follows:
H1a. Leader-group prototypicality has a positive impact on perceived leader competence.
When new group leaders enter a group, they take the position of a newcomer and therefore have a different starting point than the group leader of a newly formed group. The acceptance of a newcomer is enhanced “when this person clearly behaves in socially assimilative ways” (Rink, Kane, Ellemers, & van der Vegt, 2013: 282). Accordingly, group members are more accepted when they are perceived to be prototypical to the group and might even be included as a member of the in-group. The perceived willingness of the newcomer to identify with group norms is an important reference point in this matter (Joardar, Kostova, & Ravlin, 2007).
A positive effect of group newcomers results in the enhancement of reflection processes of previous group members and their willingness to improve task behaviors. The performance of the group after the entrance of a newcomer might improve correspondingly (Rink et al., 2013). Additionally, performance is of crucial importance for the acceptance of newcomers, as they are more likely to be accepted if the group perceives their presence as value adding to the group. In turn, group members are more willing to embrace the newcomer as a full group member (Joardar et al., 2007). This relates to the previously elaborated phenomenon that non-prototypical group leaders have to prove their intentions to the group before being accepted. Thus, newcomers may improve group performance by increasing reflection processes in the group and by adding valuable new knowledge.
New group members are generally perceived to be more non-prototypical in relation to the other group members. Nevertheless, they may be valued for their task-related input if the
group expects a performance improvement by their contribution, even if they are not accepted as members of the in-group (Joardar et al., 2007; Okimoto & Wrzesniewski, 2012). It is therefore possible that leaders’ competence is perceived to be high, although they are regarded as being non-prototypical compared to the rest of the group. As Okimoto and Wrzesniewski (2012) found out, group members that regard themselves as non-prototypical show more task effort and more self-presentation effort than prototypical group members.
Accordingly, non-prototypical group members may contribute highly to work related issues.
Once the new group leader has proven that he or she contributes to group performance and acts in favor of the group, even a non-prototypical group leader may prove his competence to the group members and receive acceptance.
Group members are more likely to accept their leader when they perceive him or her to provide a valuable contribution to the group performance. One indicator for the leaders’
expected performance is their perceived effectiveness and success. Accordingly, when the leader is perceived to be more successful he or she is more easily accepted by his or her subordinates. However, for this process to be successful group members need to have the opportunity to observe the leader’s previous behavior (Javidan, Bemmels, Stratton Devine, &
Dastmalchian, 1995). This results in the next hypothesis:
H1b. Perceived leader competence positively influences leadership acceptance.
A combination of the first two hypotheses shows a mediating relationship between the three variables.
H1c. The positive relationship between leader-group prototypicality and leadership acceptance is mediated by perceived competence of the leader.
Previous Group Performance
Prototypical group members inherit beliefs and perceptions that represent the ideal group member and are therefore perceived to act more in favor of the group. As group members identify more with a prototypical group leader than with a non-prototypical leader, the more prototypical an individual is perceived to be, the more likely it is that he or she is regarded as a member of the in-group (Joardar et al., 2007). Non-prototypical group leaders are evaluated more critically, as they are categorized as being part of the out-group. This effect enhances the degree to which the in-group regards itself as superior to the out-group.
The in-group bias therefore becomes stronger depending on previous achievements of the group in relation to the performance of the out-group (Tajfel & Turner, 2004).
Giessner, van Knippenberg and Sleebos (2009) studied how evaluations of prototypical and non-prototypical group leaders differ after poor group performance. They found that prototypical group leaders are evaluated more positively after failing to reach a performance goal than non-prototypical group leaders, which is referred to as the “license to fail” phenomenon. Non-prototypical group leaders have to prove more than prototypical group leaders that they are willing to act according to the groups’ interests. Thus, after a non- prototypical leader has succeeded in reaching a performance goal, the group regards him or her as more effective than before. Nevertheless, after failing a performance goal the prototypical leader is still perceived to be more effective than the non-prototypical leader.
This is the case because prototypical leaders receive more trust from their group members (Giessner, van Knippeberg, & Sleebos, 2009). The prototypical leader therefore has the advantage of a more positive evaluation of his or her behavior compared to a non-prototypical leader. However, this effect could change after a leadership change. As group members would have the opportunity to differentiate themselves from their previous group leader who showed poor performance, the license to fail effect would not be necessary for group members to
preserve their positive group image. Instead, they might neglect the identity they had developed under the previous group leader and form a new and more positive identity with the new group leader.
With the entrance of a new group leader, each group member forms an anticipation of how that new leader will behave. This anticipation is based on previous experiences and influences the perceived competence of and satisfaction with the group leader. Given that the performance expectation of a prototypical individual is high, that individual will get more opportunities to prove his capabilities in the group-context as he or she is granted a greater amount of attention and influence. According to Expectation States Theory, a person is more satisfied with other group members when their initial expectations about that person are met (Correll & Ridgeway, 2006). As prototypical group leaders are initially evaluated more positively and therefore create high expectations, in case of low task performance these expectations are disappointed drastically. Therefore, we propose that if the new group leader is prototypical as well, he or she is evaluated based on the disappointed expectations experienced earlier. A non-prototypical new group leader would not carry this stigmatization.
Task failure may lead to a more negative self-identity of a person or group.
Prototypical group leaders, as they are regarded as part of the in-group, are likewise evaluated with this negative image. The trust in their competence decreases and the perceived competence of out-group members increases. Newcomers are less likely to be perceived prototypical but they may still be valued for their task-related competencies without being accepted as part of the in-group (Joardar et al., 2007; Okimoto & Wrzesniewski, 2012).
Therefore, if the group has performed poorly with a prototypical group leader and this person is associated with a negative image due to the low performance, the task-related input of a new non-prototypical leader might be valued more positively than the input of a new prototypical leader; as a non-prototypical new leader does not carry this negative image (Schippers, Homan, & van Knippenberg, 2013). Furthermore, because failure in goal
Leader-Group Prototypicality
Perceived Leader Competence
Leadership Acceptance Previous
Performance
+ +
+
achievement increases team reflexivity, and reflexivity is positively related to team performance, the combination of a new group leader and poor prior performance may affect the group members’ preference of group leader characteristics. As the non-prototypical new group leader is not subject to the negative attribution of the previous group leader, it is likely that group members are more willing to consider contributions and to accept the influence of non-prototypical group leaders. Accordingly, we suggest that a non-prototypical group leader will be evaluated more positively after the failure of a prototypical group leader.
H2. The relation between prototypicality and perceived competence will be stronger and positive when previous performance is high, than when it is low.
The hypotheses elaborated above result in a conditional indirect effect (Figure 1). This concludes:
H3. The indirect effect of leader prototypicality on leader acceptance through perceived leader competence is stronger when previous performance is high but weaker when previous performance is low.
H1a H2
H1b H1c H3
FIGURE 1 Conceptual Model
METHOD
Participants and Design
To test the hypotheses outlined above, a laboratory experiment was performed. The experiment was conducted with 84 students of the University of Groningen. Of the final participants 77.1% were female, and 29.9% were male. The mean age was 22.44 years.
Participating students came from a range of study fields covering social sciences, natural sciences, physical sciences, and applied sciences programs.
The experimental groups were divided in a 2x2 design, manipulating leader prototypicality and task performance. Each group consisted of two group members and one group leader that were randomly selected. Because we were interested in perceptions of leader competence and leader acceptance of group members, only the answers of group members were used for the analyses. The number of participants in each condition is presented in Table 1. The total number of participants for the analyses was 56.
TABLE 1 Experimental Conditions
Manipulation of Task Performance
Manipulation of Prototypicality
Leader-Group Total
High Low
High
12 16 28
Low 12 16 28
Total 24 32 56
Procedure and Manipulations
Each session was conducted with six participants. They came to a room that was equipped with six chairs around one table, each marked with a letter A-F to identify the participants. Each participant received an informed consent form and an experiment guideline
that specified session procedure and special tasks of the group leaders. The special tasks of the group leaders included guiding the discussions, ensuring equal contribution to the solution of every group member, motivating team members, and making sure that the group completed the task within the given time frame (see appendix A for experiment guideline). The participants could freely choose their seats. In the first part of the experiment they completed a “brain hemisphere test”1 consisting of 16 items and were informed that the final group composition would be based on the outcome of this test.
Participants A, B, and E were then asked to move to a second experimental room.
Participants C, D, and F stayed in the first room. Participants within each of the subgroups received feedback that they had scored similarly on the brain hemisphere test and were therefore assigned to the same subgroup. Their supposed outcomes of the test were presented graphically (see appendix B). Participants B in one group and D in the other were marked to have scored in the middle of their three person group and were therefore assigned to be the group leader. The groups then received their first group task which involved a survival scenario. In this scenario the group members were lost on the Atlantic Ocean as the yacht they were on had sunk. The members could save 15 items of the sinking ship and had to rank order those items from the most important (1) to the least important (15). The group would use the five top ranked items for their survival (Graham Knox, Lost at Sea2). Each group had a time limit of 10 minutes to complete the task.
After the time had passed, task solutions were collected and the participants were asked to sit in different corners of their rooms to complete the first questionnaire, in which we measured perceived prototypicality of group leaders in relation to the group members, perceived competence of group leaders, leadership acceptance, satisfaction with the group, satisfaction with the leader, and amount of group conflict. After the answers for the first questionnaire were collected, group leaders were asked to leave the room. The remaining
1 http://www.ipn.at/ipn.asp?BHX
2 http://insight.typepad.co.uk/lost_at_sea.pdf
group members then received the manipulated task feedback and were informed about the scores on the brain hemisphere test of the other group leader that would enter their subgroup for the next task. The task feedback was provided on a scale from 0 (very good) to 71 (very poor). The average score for this task was 33-45. Results were supposedly calculated by adding up the deviations between the group ranking and the ranking of the optimal task solution. Participants in the high task performance condition were told to have scored 17 points, while participants in the low task performance condition were told to have scored 61 points. Feedback was provided graphically using a scale that showed the average score and the score of the group (see appendix C). Rating procedures were explained to participants orally.
Depending on the experimental condition the new group leader was either prototypical to the other group members or not. Group members received information about the “brain hemisphere test” scores of the new leader verbally. The score was also marked on the same scale as the other group members’ test scores to illustrate it graphically. In the high prototypicality condition the new leader had supposedly scored in the middle of the other group members, close to the previous group leader. In the low prototypicality condition the new group leader had scored on the other side of the scale. Afterwards, participants completed a second questionnaire referring to their expectations about the new group leader with regard to his or her score on the brain hemisphere test. This questionnaire measured perceived prototypicality, perceived competence, and leadership acceptance.
Group members did not receive any information about the performance of the other experimental subgroup or the other group leader. After completing the second questionnaire, the new group leader entered the room and the group was provided with the second task, for which they again had a time limit of 10 minutes. This task again consisted of a survival scenario. In the scenario the group had survived a plane crash in northern Canada during winter. The group had 12 items which they had to rank order from the most important (1) to
the least important (12) and would use the five top tanked items for their survival3. Participants did not receive information about their performance on the second group task.
The experiences with the new group leader were measures with a third questionnaire.
Measures
The first and third questionnaire measured the variables perceived prototypicality of group leaders in relation to the group members, perceived competence of group leaders, leadership acceptance, satisfaction with the group, satisfaction with the leader, and amount of group conflict. The second questionnaire only measured the first three variables. Each variable was measured using 5-point Likert scales (1=disagree to 5=agree). Group conflict was measured on a rating from 1=none to 5=a lot. The wording of the questions at the different measurement times differed based on the reference to the old or new group leader and whether it referred to the actual experiences with the leader (questionnaire 1 and 3) or their expectations about them (questionnaire 2).
The variable perceived prototypicality was measured using six items. The questions examined how group members viewed the group leaders based on their similarities to the other group members, stating “overall, I would say that the leader represents what is characteristic about group members of this group” or “overall, I would say that the leader is very similar to most group members” (c.f. Platow & van Knippenberg, 2001). The reliability of this variable was measured using the Cronbach’s Alpha, showing high reliability at all measurement times: =.860 in the first questionnaire, =.898 in the second questionnaire, and
=.861 in the last questionnaire.
Items used to measure perceived leader competence included “the group leader provided a valuable contribution to our group” and “I trusted the group leader with our task”.
3 http://scoutingweb.com/scoutingweb/SubPages/SurvivalGame.htm
This was used to assess the perceived general competence of the newly introduced group leader. The Cronbach’s Alpha for this variable is =.883, .707, and .923 for questionnaire 1, 2 and 3 respectively.
Leadership acceptance was measured by a 10-item questionnaire with a similar scale.
Acceptance was measured as both task-based acceptance as indicated by the first five items requiring an evaluation in terms of “the group leader was an asset to our workgroup”, and relationship-based acceptance which is measured by the last five items. Relationship-based acceptance was assessed using questions similar to “the group leader was socially compatible with our group” (c.f. Joardar et al., 2007). The questionnaires for leadership acceptance were reliable: =.945, .849, and .940, for questionnaire 1, 2 and 3 respectively.
Additional variables assessed in the first and third questionnaire were satisfaction with the group, satisfaction with the leader and amount of perceived conflict. These were not included in the hypotheses, however, they were included in the questionnaires as they could provide valuable additional information.
Satisfaction of the team members was measured in terms of satisfaction with the group, consisting of seven items, and satisfaction with the leader, consisting of five items.
Items referring to satisfaction with the group considered group atmosphere such as “I felt comfortable with my group”, and group collaboration such as “all group members provided good and useful ideas” (c.f. Keyton, 1991). Cronbach’s Alpha was =.856 in the first questionnaire, and =.908 in the third questionnaire. Satisfaction with the group leader was likewise assessed with items referring to social satisfaction, including the statement “I felt comfortable with the group leader” and in terms of the leaders fulfillment of his role as specified in the experiment guidelines, such as “The group leader handled disagreements during the group task well”. Reliability for this scale was =.865, and =.907 for questionnaires 1 and 3 respectively.
Group conflict was measured using a 5-point Likert scale ranging from 1=none to 5=a lot. The items assessed conflict in decision making during the task solution process in the first and second task. Items used included “How frequently were there conflicts about ideas in your group?” (c.f. Jehn, 1995). The reliability was =.871 in the first questionnaire, and
=.884 in the third questionnaire.
RESULTS
Questionnaire 3 measured the participant’s perception of characteristics of the new leader after the second group task and therefore provided results with less speculation than the results of questionnaire 2 because it was based on experience rather than expectations of the participants. As we were interested in effects on an individual’s perception, results from questionnaire 3 are presented below. The analysis was conducted with the results from questionnaire 2 as well, referring to expectations about the new leader, leading to similar or less significant results.
Manipulation Check
A manipulation check for leader-group prototypicality was conducted in the second questionnaire, to see whether the manipulation on the perception of the leader’s prototypicality for group members was successful. A 2x2 ANOVA was conducted using manipulated prototypicality and manipulated task performance as independent variables and perceived prototypicality during the second questionnaire as dependent variable.
TABLE 2 Manipulation Check
Manipulation of Task Performance
Manipulation of Prototypicality Leader-Group
Mean Standard
Deviation N
High yes 3.49 .72 12
no 2.97 .83 16
Low yes 3.69 .58 12
no 3.42 1.05 16
Total yes 3.59 .65 24
no 3.19 .96 32
The results in table 2 show that the manipulation did result in lower perceived prototypicality for the low prototypicality condition (M=3.19) than for the high prototypicality condition (M=3.59). However, the main effect of prototypicality has proven to be insignificant (F(1,52)=3.105, p=.084). The attempted manipulation of leader-group prototypicality therefore seems to not have been successful. No manipulation check was conducted for the manipulation of task performance.
Descriptive statistics of the third questionnaire are presented in table 3. The correlations among all variables are highly significant and show strong connections.
Perceived group conflict is the only variable showing a negative correlation to the others, suggesting that there is perceived to be less group conflict if the group leader is more prototypical, more competent, and more accepted by the group members. The results show the strongest correlation between perceived competence and satisfaction with the leader, indicating that the more competent the leader is perceived to be, the more satisfied group members are with him or her.
TABLE 3 Descriptive Statistics
Measure Mean SD 1 2 3 4 5 6
1 Prototypicality 3.34 .79 1 2 Perceived
Competence 3.69 .88 .68* 1
3 Leadership
Acceptance 3.54 .85 .76* .75* 1
4 Satisfaction
with group 4.07 .75 .71* .75* .74* 1
5 Satisfaction
with leader 3.73 .91 .65* .84* .77* .75* 1
6 Group
Conflict 2.48 .83 -.59* -.43* -.51* -.53* -.39* 1
* Correlation is significant at the 0.01 level (2-tailed)
Hypotheses Testing
Although the manipulation check for leader-group prototypicality suggests that the manipulation was not successful, the hypothesis will be tested using the manipulated prototypicality as the independent variable to identify any possible manipulation effects.
To test the hypotheses 1a to 1c a regression analysis was conducted using the regression model 4 of the PROCESS add-on by Andrew F. Hayes. The analysis includes manipulated prototypicality as independent variable, leadership acceptance as dependent variable, and perceived competence as the mediator to this relationship. Hypothesis 1a predicted that higher manipulated prototypicality leads to higher perceived competence. The results show a slight positive effect for this relationship (b=.13) but this effect was not significant (t=.53, p>.05). Therefore, hypothesis 1a was not supported. Hypothesis 1b suggested that perceived competence positively affects leadership acceptance. The results show a strong positive effect (b=.73). The relation is highly significant (t=8.24, p<.005).
Hypothesis 1b was supported.
Hypothesis 1c predicted that the positive relationship between manipulated prototypicality and leadership acceptance is mediated by perceived competence. The analysis showed a small direct effect of leader-group prototypicality on leadership acceptance (b=.10) and an indirect effect mediated by perceived competence (b=.09). However, the mediation was not significant (p>.05, 95% CI [-.21, .41]). Hypothesis 1c was therefore not supported.
Hypothesis 2 and 3 imply a moderated mediation analysis which was conducted with model 7 of the PROCESS approach. The results of this analysis are presented in Table 4. A moderated mediation model was used applying manipulated prototypicality as independent variable, leadership acceptance as dependent variable, perceived competence as mediator and manipulated task performance as a first-stage moderator. The analysis showed a negative interaction effect of manipulated prototypicality and manipulated task performance on
perceived competence (b=-.32) but the results were not significant (t=-.66, p>.05). Thus, hypothesis 2 was not supported.
TABLE 4
Effects of Moderated Mediation Analysis (Independent Variable: Manipulated Prototypicality)
Note: * p>0.1, ** p< 0.0005
Hypothesis 3 predicted that the indirect effect of leader prototypicality on leadership acceptance through perceived competence would be stronger when previous performance was high than when it was low. The analysis showed a positive mediated effect in the case of high task performance and a slightly negative effect in case of low task performance (see Table 5).
However, mediation was not significant, neither when task performance was high, nor when it was low (p>.05). The results suggest that hypothesis 3 was not supported.
Independent Variable
Dependent Variable
Perceived Competence Leadership Acceptance
Coefficient t Coefficient t
Manipulated
Prototypicality .29* .83 .10* .67
Manipulated Task
Performance .24* .65 - -
Interaction -.32* -.66 - -
Perceived
Competence - - .72** 8.24
Conditional Indirect
Effects Effect Bootstrap Confidence Interval
Lower Level Upper Level
Task Performance
high .21 -.18 .66
Task Performance
low -.02 -.54 .61
Additional Analysis
As the manipulation of leader-group prototypicality seemed to not have been successful, additional analyses using measured perceived prototypicality instead of manipulated prototypicality were conducted.
Similar to the previous analysis, hypotheses 1a and 1c, with the independent variable perceived prototypicality, were analyzed using model 4 of the PROCESS approach by Hayes.
Hypothesis 1a predicted a positive relationship between perceived prototypicality and perceived competence of the leader. The analysis showed a strong relation between perceived prototypicality and perceived competence (b=.75, t=6.74, p<.005). Therefore, results suggest that higher perceived prototypicality does indeed lead to higher perceived competence of the leader. Hypothesis 1a was supported. As hypothesis 1b was already tested and supported above with the same variables, this analysis will not be repeated.
Hypothesis 1c predicted that the relationship of perceived prototypicality on leadership acceptance was mediated by perceived competence. The analysis showed a positive indirect effect (b=.31) of perceived prototypicality on leadership acceptance. The model proved to be highly significant (p<.005, 95% CI [.14, .56]). The results suggest that high perceived prototypicality does lead to increased leadership acceptance and that this relationship is mediated by perceived leader competence. Furthermore, a direct effect of perceived prototypicality on leadership acceptance was found (b=.50, t=4.47, p<.005). Hence, hypothesis 1c was supported for perceived rather than manipulated prototypicality.
The analysis of hypotheses 2 and 3 was conducted using PROCESS model 7. Table 5 shows the results of this analysis. Hypothesis 2 predicted that the relationship between perceived prototypicality and perceived competence would be less positive in the low task performance condition than in the high task performance condition. The results show a positive interaction effect of perceived prototypicality and task performance manipulation on
perceived competence (b=.36). This suggests that unlike hypothesized, the relationship between perceived prototypicality and perceived competence was stronger under high rather than low task performance. Hypothesis 2 was therefore not supported. The interaction proved to be insignificant (t=1.56, p>.12), however, results indicate a tendency of the interaction effect.
TABLE 5
Effects of Moderated Mediation Analysis (Independent Variable: Perceived Prototypicality)
Independent Variable
Dependent Variable
Perceived Competence Leadership Acceptance
Coefficient t Coefficient t
Perceived
Prototypicality .54** 3.02 .50*** 4.47
Manipulated Task
Performance -1.20* -1.55 - -
Interaction .36* 1.56 - -
Perceived
Competence - - .42*** 4.14
Conditional Indirect
Effects Effect Bootstrap Confidence Interval
Lower Level Upper Level
Task Performance
high .22 .06 .47
Task Performance
low .37 .14 .68
Note: * p>0.1, ** p<0.005, *** p<0.0005
Hypothesis 3 predicted that the indirect effect of perceived prototypicality and task performance through perceived competence would be stronger under the condition on high task performance than under low task performance. However, the results showed a stronger indirect effect for low rather than high task performance. The results indicate that participants in the low task performance condition evaluated a prototypical group leader more positive than participants in the condition of high task performance. Contrary to the prediction of H3 the results indicate that the effects of task performance on leadership acceptance are stronger under the condition of low task performance, rather than high task performance. The model
does not show a significant interaction effect of task performance on perceived competence.
The tendency of this interaction effect, however, is negative.
DISCUSSION
The objective of this study was to identify how prior task performance affects the groups’ preferences for prototypical leaders. We argued that in the case of low group performance, a non-prototypical group leader would be preferred after a leadership change.
Thus, we predicted that low group performance would lead to a more positive evaluation of non-prototypical group leaders in terms of competence and leadership acceptance. Hypotheses based on manipulated new leader prototypicality were not supported. Because the manipulation check raised doubts about success of the prototypicality manipulation we conducted the same analysis using the participants perceived rather than manipulated new leader prototypicality. The results indicated that perceived prototypicality positively affected perceived competence of the new leader and leadership acceptance. This effect was stronger in the low task performance condition than in the high task performance condition.
The manipulation of prototypicality in this study was conducted using a brain hemisphere test based on previous studies (for example see van Knippenberg & van Knippenberg, 2005). However, results of the conducted analyses suggest that the manipulations in our experiment were not successful. Participants were provided with results of the brain hemisphere test scores in terms of a graphic illustration of all group member’s scores on a scale ranging from left brain hemisphere dominated to right brain hemisphere dominated. It is possible that this provided feedback was not perceived to be credible by the participants. Furthermore, as the participants interacted with each other during the group tasks after having received the feedback, it is likely that their experiences during the collaboration were more influential in creating an assessment of prototypicality of group members than the feedback of the brain hemisphere test scores. This would also explain why the analyses conducted with the results for perceived prototypicality from questionnaire 2 and 3 provided more significant results.
In line with previous findings, the analyses showed a strong relationship between prototypicality and leader acceptance (Hains, Hogg, & Duck, 1997). Additionally, we could provide support for a mediation effect of perceived competence. The analysis showed that the three variables are strongly interrelated and a mediating effect of perceived competence of the group leader was found (H1a and H1c). Likewise, a strong effect of perceived competence on leadership acceptance was found (H1b). Contrary to the theoretical reasoning in this paper, the results suggest that leaders with high perceived prototypicality are more positively evaluated under conditions of low previous task performance rather than high previous task performance (H2 and H3). However, these results can only be referred to as trends, as effect levels do not reach significance.
Possible reasons for findings contrary to our hypothesis can be found in Social Identity Maintenance Theory (Turner & Pratkanis, 1998). Group members try to maintain their positive group-image or try to find ways to enhance the group-image if this is threatened.
When low task performance has threatened the positive self-image of a group, members are likely to try to attribute their low performance to influences they had no control over. The reason for low task performance is therefore not attributed to the leader or his or her characteristics but to outside influences. Accordingly, the collective identity of the group is not threatened as the positive image of the group leader is not decreased. The group focuses on aspects or characteristics that help to sustain a collective identity instead of the differences between the group members. This effect is especially strong when the group identity is threatened, for example as in this case by low task performance (Turner & Pratkanis, 1998).
Therefore, instead of developing a less positive expectation of the group leader based on prototypicality, the group members increased their positive group-image. Accordingly, as other external factors were attributed to be responsible for poor group performance, the group members’ trust in prototypical group leaders did not decrease and therefore their evaluation of the new group leaders’ competence and their leadership acceptance was not affected.
As elaborated before, the expectations of group members about their new leader are of crucial importance for their future acceptance and satisfaction with this leader. The expectations of each group member were assessed using questionnaire 2. An additional analysis was conducted aiming at identifying the effect of previous expectations about the new leader on the relationship between perceived prototypicality and leadership acceptance in the third questionnaire. However, the results did not show any significant effects.
Nevertheless, we assume that the expectations of group members influenced their subsequent assessment of the leaders.
We predicted that to retain a positive group-image group members would prefer a non- prototypical group leader after having performed poorly, because they would expect a performance improvement which would then enhance the self-image of the members.
However, the process of self-image preservation seems to be directed at a different process than expected. It seems that the group members instead tried to preserve their positive group- image by searching for explanations by external influences for their low task performance.
The main goal of the group in this case was not to improve their subsequent task performance but instead to protect their positive group identity (Turner & Pratkanis, 1998). Accordingly, as low task performance was not attributed to the previous group leader, the self-image of the group was not decreased, because the image of the leader did not suffer from the poor performance. This also relates to the principle of ‘license to fail’ also referred to by Giessner et al. (2009), in which prototypical group leaders are more positively evaluated than non- prototypical group leaders in cases of low performance.
As Ellemers, de Gilder and Haslam (2004) have shown, the leaders’ identity on which group members focus depends on situational features. Accordingly, members either focus on aspects that differentiate the leader from the members or on aspects that are similar between them, and this influences their leader acceptance. After the group members have interacted with the group leaders and had the opportunity to gain a more thorough understanding of their
characteristics it is possible that the group members adapted their initial assessment of whether the group leader was prototypical or not. As group members tend to focus on aspects that emphasize the similarities of group members in order to create a cohesive group image (Turner & Pratkanis, 1998), they possibly altered their perceived prototypicality of a leader when they were more satisfied with the leader’s performance, thereby refusing the prototypicality of their previous group leader with whom they had performed poorly and enhancing the prototypicality of the new group leader. It is possible that the results deviated from our predictions as non-prototypical group leaders that show high performance are evaluated as being more prototypical because this is more beneficial for a positive group- image. This may be the case if the leaders’ behavior is perceived to be desirable for the group goals and whether it matches the shared group identity (Ellemers et al., 2004).
The willingness and effort of a new group member to adapt to existing group norms is an important reference point to his or her acceptance and perceived prototypicality by other group members (Joardar et al., 2007). We did not analyze any group processes during the group collaboration phases of this study and therefore are not able to judge the interactions between group members. It is possible that new group leaders adapted to existing group norms and therefore influenced their perceived prototypicality in the new group. As it is especially important for the acceptance of non-prototypical group leaders that they are perceived to identify with the other group members (Steffens, Schuh, Haslam, Pérez, & van Dick, 2015), it is possible that this process limited the effectiveness of our prototypicality manipulation.
Individuals occupy different identities and these identities are context dependent.
Different identities become salient depending on the situations and expectations they have to meet (Stryker & Burke, 2000). It is possible that conflicts develop about which identity becomes more salient in the respective situation (Stets & Burke, 2000). In this case it seemed to be more important to the new leaders to adapt to the group and therefore create a new
group identity, than to adhere to the role identity of their leader position which would make them more of an out-group member. Potentially, the group-identity of the leaders therefore became more salient than their role identity, making them appear more prototypical.
Implications
Although some findings of this study did not show significant results regarding the moderating role of task performance on the relationship between perceived prototypicality and leadership acceptance, the study provides useful insights into the current literature and a basis for future research directions.
Results of this study suggest a combination of research on group prototypicality and Social Identity Theory. Prototypicality may not only have an influence on whether an individual is perceived to be an in-group or out-group member, but may also affect which identity is most important to an individual in which situation. Based on the findings of this study, we hypothesize that group-identity may be especially important for non-prototypical group members, while role-identity or personal-identity would become more salient for leaders that are prototypical to the other group members. A combination of those two theories therefore appears to be highly beneficial.
Furthermore, the findings add to the line of prototypicality research by adding that perceived competence positively mediates the relationship between prototypicality and leader acceptance. The influence of task performance in this matter is not clear, as findings were not significant. The role of task performance on the relationships among prototypicality, perceived competence and leadership acceptance therefore still needs to be researched and will provide further insights. An experiment providing realistic task feedback that is not subject to manipulation and a task where participants can more easily assess their own performance is likely to enhance the significance of these results.
Findings of this study provide further insight to group newcomer research. In our study the new group leader occupied the position of the new group member. Acceptance of the new leader was not only dependent on his or her status as a new group member but also on the previous performance the group members experienced. Combined with Social Identity Theory it becomes likely that not only the possible contribution of the new leader to group performance but also identity motives of group members are important determinants for group newcomer processes.
Implications for practice can be derived because the relationship between group leaders and group members is of importance for the collaboration process and the output quality of a group. Especially in professional situations where a leader is appointed and does not evolve, the choice of a leader should be based on careful considerations. Results of this study suggest that group members will prefer a prototypical group leader to a non-prototypical group leader, even after having experienced poor performance with such prototypical leaders.
Therefore, especially in occupations and business areas where group leaders or supervisors change regularly, it is highly important to not only select a leader based on his qualification and personal characteristics, but also to evaluate his or her suitability in relation to colleagues and subordinates.
Limitations and Future Research
The manipulation of prototypicality in our study was based on similar manipulations successfully conducted in studies by van Knippenberg and van Knippenberg (2005).
However, in our experiment the manipulation resulted in only weak effects. The assessment of a person’s prototypicality is likely to be based on multiple individual characteristics and subject to a complex interaction of different influences. After having interacted with the other group members, each participant has created a more thorough perception of the other
participants’ characteristics. It is therefore possible that our manipulation of prototypicality was not sufficient to change the participants’ perception of the group leaders’ prototypicality with regard to the group. Furthermore, as the group prototype represents the ideal group member, it is likely that a manipulation in terms of a brain hemisphere test did not provide an adequate reference point for the group prototype. Identifying which traits influence the perceived prototypicality and to what extent would greatly benefit the literature in this field.
Due to the limited time frame available for our study, the number of total participants for the study was 84. This sample size limits the statistical power of the findings, however, results indicate that an indirect mediation effect of task performance through perceived prototypicality and perceived competence on leadership acceptance does exist and may be further examined in broader studies with more participants and a longer collaboration timeframe.
We suggest that the new group leaders actively tried to adapt to group norms in order to create a group identity and thereby possibly neglected their role identity as group leader or their individual identity. An interesting outlook for future research is to analyze the hierarchy different identities take for an individual and under which conditions these become salient.
The findings suggest that in this case the individual decided that common group identity would be more important or beneficial to him or her, than role identity or individual identity would. Circumstances and premises that are responsible for one identity becoming more dominant or salient than the other leave room for future research.
All assessed variables show strong correlations with each other (see Table 3). This may indicate that the participants had difficulties separating the assessment of the different variables in the questionnaire or that the items did not differentiate the constructs sufficiently.
Although the assessment of most variables was based on previous studies, their combination as applied in this study may have not been optimal. For future research it is therefore
advisable to separate the constructs more strictly or to inform the participants in advance about the definitions of the items.
Likewise, effects of expectations about the new leader on the subsequent assessments could not be found in this study. The effect of expectation states theory therefore cannot be identified in this study but should be researched in future studies. A more profound personality feedback of the new group leader before entering the group would provide the participants with a better basis for their expectations. A detailed assessment of these expectations would benefit this line of research by providing more qualitative results.
This study was conducted with university students in a controlled environment.
Expanding the research to organizational work groups could complement the gained insights.
Resulting from the limited experimental time-frame it was not possible to identify what the groups based their perceptions of prototypicality on. This limitation could be overcome using a longitudinal study in which group members and leaders collaborate and get to know each other over a longer time period. Further studies could contribute to these findings by identifying which situational variables influence which identity becomes salient in a specific situation and under which circumstances group members choose to neglect a role identity instead of trying to preserve it.
The amount of time the group members have spent with each group leader could further influence the outcome of the experiment. In our experiment the groups had a fixed timeframe of 10 minutes time for each task. During this limited time period only a superficial understanding of each other’s characteristics and the similarities could develop. An interesting aspect for future research would therefore be to increase the amount of time the groups get to socialize with each other. This would also create the opportunity to base the analysis on actual prototypicality of the participants and not on manipulated prototypicality.
CONCLUSION
Our findings provided support for the mediation effect of perceived leader competence on the relationship between prototypicality and leadership acceptance. The study could not provide significant results for the effect of task performance in this matter. The results, however, suggest that contrary to initial reasoning the processes of preserving the positive group image of participants is not directed towards the creation of a more positive image under the new leader, but instead is focused on preserving and enhancing the group image that has been created under the previous group leader.
REFERENCES
Cicero, L.; Pierro, A.; van Knippenberg, D. 2009. Leadership and Uncertainty: How Role Ambiguity Affects the Relationship between Leader Group Prototypicality and Leadership Effectiveness. British Journal of Management, 21: 411–421.
Correll, S. J.; Ridgeway, C. L. 2006. Expectation States Theory. In John Delamater (Ed.), Handbook of Social Psychology: 29–51: Springer US.
Ellemers, N.; De Gilder, D.; Haslam, S. A. 2004. Motivating Individuals and Groups at Work: A Social Identity Perspective on Leadership and Group Performance. Academy of Management Review, 29(3): 459–478.
Giessner, S. R.; van Knippenberg, D.; Sleebos, E. 2009. License to fail?: How leader group prototypicality moderates the effects of leader performance on perceptions of leadership effectiveness. The Leadership Quarterly, 20(3): 434–451.
Hains, S. C.; Hogg, M. A.; Duck, J. M. 1997. Self-Categorization and Leadership:
Effects of Group Prototypicality and Leader Stereotypicality. Personality and Social Psychology Bulletin, 23(10): 1087–1099.
Hogg, M. A. 2001. A Social Identity Theory of Leadership. Personality and Social Psychology Review, 5(3): 184–200.
Hogg, M. A.; Terry, D. J. 2000. Social Identity and Self-Categorization Processes in Organizational Contexts. Academy of Management Review, 25(1): 121–140.
Hogg, M. A.; van Knippenberg, D. 2003. Social Identity and Leadership Processes in Groups. Advances in Experimental Social Psychology, 35: 1–52.
Javidan, M. 1995. Superior and Subordinate Gender and the Acceptance of Superiors as Role Models. Human Relations, 48(11): 1271–1284.
Jehn, K. A. 1995. A Multimethod Examination of the Benefits and Detriments of Intragroup Conflict. Administrative Science Quarterly, 40(2): p. 256.
Joardar, A.; Kostova, T.; Ravlin, E. C. 2007. An experimental study of the acceptance of a foreign newcomer into a workgroup. Journal of International Management, 13(4): 513–
537.
Kahai, S. S.; Sosik, J. J.; Avolio, B. J. 1997. Effects of Leadership Style and Problem Structure on Work Group Processes and Outcomes in an Electronic Meeting System Environment. Personnel Psychology, 50: 121–146.
Kahai, S. S.; Sosik, J. J.; Avolio, B. J. 2003. Effects of leadership style, anonymity, and rewards on creativity-relevant processes and outcomes in an electronic meeting system context. The Leadership Quarterly, 14(4-5): 499–524.
Keyton, J. 1991. Evaluating Individual Group Member Satisfaction as a Situational Variable. Small Group Research, 22(2): 200–219.
Okimoto, T. G.; Wrzesniewski, A. 2012. Effort in the face of difference: Feeling like a non-prototypical group member motivates effort. European Journal of Social Psychology, 42(5): 628–639.
Pierro, A.; Cicero, L.; Bonaiuto, M.; van Knippenberg, D.; Kruglanski, A. W. 2005.
Leader group prototypicality and leadership effectiveness: The moderating role of need for cognitive closure. The Leadership Quarterly, 16(4): 503–516.
Platow, M. J.; van Knippenberg, D. 2001. A Social Identity Analysis of Leadership Endorsement: The Effects of Leader Ingroup Prototypicality and Distributive Intergroup Fairness. Personality and Social Psychology Bulletin, 27(11): 1508–1519.
Rink, F.; Ellemers, N. 2008. Diversity, newcomers, and team innovation: The importance of a common identity. In, Diversity and Groups, vol. 11: 221–243. Bingley:
Emerald (MCB UP ).
Rink, F.; Kane, A. A.; Ellemers, N.; van der Vegt, G. 2013. Team Receptivity to Newcomers: Five Decades of Evidence and Future Research Themes. The Academy of Management Annals, 7(1): 247–293.
Schippers, M. C.; Homan, A. C.; van Knippenberg, D. 2013. To reflect or not to reflect: Prior team performance as a boundary condition of the effects of reflexivity on learning and final team performance. Journal of Organizational Behavior, 34(1): 6–23.
Sicasubramaniam, N.; Murry, W. D.; Avolio, B. J.; Jung, D. I. 2002. A Longitudinal Model of the Effects of Team Leadership and Group Potency on Group Performance. Group
& Organization Management, 27(1): 66–96.
Steffens, N. K.; Schuh, S. C.; Haslam, S. A.; Pérez, A.; van Dick, R. 2015. ‘Of the group’ and ‘for the group’: How followership is shaped by leaders' prototypicality and group identification. European Journal of Social Psychology, 45(2): 180–190.
Stets, J. E.; Burke, P. J. 2000. Identity Theory and Social Identity Theory. Social Psychology Quarterly, 63(3): 224–237.
Stryker, S.; Burke, P. J. 2000. The Past, Present, and Future of an Identity Theory.
Social Psychology Quarterly, 63(4): 284–297.
Tajfel, H.; Turner, J. C. 2004. The Social Identity Theory of Intergroup Behavior. In John T. Jost, Jim Sidanius (Eds.), Political psychology: 7–24. New York: Psychology Press.
Tanghe, J.; Wisse, B.; van der Flier, H. 2010. The Formation of Group Affect and Team Effectiveness: The Moderating Role of Identification. British Journal of Management, 21: 340–358.
Turner, M. E.; Pratkanis, A. R. 1998. A Social Identity Maintenance Model of Groupthink. Organizational Behavior and Human Decision Processes, 73(2/3): 210–235.
van Knippenberg, B.; van Knippenberg, D. 2005. Leader Self-Sacrifice and Leadership Effectiveness: The Moderating Role of Leader Prototypicality. Journal of Applied Psychology, 90(1): 25–37.
van Knippenberg, D. 2011. Embodying who we are: Leader group prototypicality and leadership effectiveness. The Leadership Quarterly, 22: 1078–1091.
van Knippenberg, D.; Hogg, M. A. 2003. A Social Identity Model of Leadership Effectiveness in Organizations. Research in Organizational Behavior, 25: 243–295.
APPENDIX A
Experiment Guideline
The purpose of this experiment is to find out how leadership change influences the group decision making process. For this purpose you will be asked to complete two short group experiments and fill out individual questionnaires in between.
In the beginning of this experiment we will ask you to conduct a test that will measure your dominant brain hemisphere. According to your test-results we will assign you to your first groups with participants that have a similar test-result as you. The results of your tests are not subject to any kind of judgment. It is not regarded as better or worse whether your left or right brain hemisphere is dominant.
One person from your group will be assigned to be the group leader. The group leaders will have to pay attention to the following tasks:
- Make sure that each group member contributes equally to the decisions made - Make sure that each group member has the opportunity to share his opinion - Guide the discussions to find a solution that every group member is satisfied with
- Keep track of the time and make sure that the task is completed within the given timeframe
- Motivate the other team members to participate in the task if necessary - Praise, support and encourage the group members for their input
In this group we will ask you to complete a group task for which you will have about 10 minutes. Afterwards we will ask everyone to fill out the first questionnaire. The group leaders will leave the observation room before we inform the remaining group members of their task scores. The group leaders will not be informed about how the groups have scored in the tasks.
For the second group task the group leaders of two participating groups will be switched. You will be either teamed up with a new group leader that has scored similar to or different from your previous group on the brain hemisphere test. Together with this new group leader you will complete a new group task. Again the timeframe will account to about 10 minutes.
