ArticlePDF Available

Leader Self-Sacrifice and Leadership Effectiveness: The Moderating Role of Leader Prototypicality.

Authors:

Abstract and Figures

Self-sacrificing behavior of the leader and the extent to which the leader is representative of the group (i.e., group prototypical) are proposed to interact to influence leadership effectiveness. The authors expected self-sacrificing leaders to be considered more effective and to be able to push subordinates to a higher performance level than non-self-sacrificing leaders, and these effects were expected to be more pronounced for less prototypical leaders than for more prototypical leaders. The results of a laboratory experiment showed that, as expected, productivity levels, effectiveness ratings, and perceived leader group-orientedness and charisma were positively affected by leader self-sacrifice, especially when leader prototypicality was low. The main results were replicated in a scenario experiment and 2 surveys.
Content may be subject to copyright.
Leader Self-Sacrifice and Leadership Effectiveness: The Moderating Role
of Leader Prototypicality
Barbara van Knippenberg
Vrije Universiteit
Daan van Knippenberg
Erasmus University Rotterdam
Self-sacrificing behavior of the leader and the extent to which the leader is representative of the group
(i.e., group prototypical) are proposed to interact to influence leadership effectiveness. The authors
expected self-sacrificing leaders to be considered more effective and to be able to push subordinates to
a higher performance level than non-self-sacrificing leaders, and these effects were expected to be more
pronounced for less prototypical leaders than for more prototypical leaders. The results of a laboratory
experiment showed that, as expected, productivity levels, effectiveness ratings, and perceived leader
group-orientedness and charisma were positively affected by leader self-sacrifice, especially when leader
prototypicality was low. The main results were replicated in a scenario experiment and 2 surveys.
For decades, practitioners and scientists have been intrigued by
those components of leadership that motivate people to higher
levels of effort and performance. Leadership researchers studied
leader traits, behavioral style, and situational contingencies to look
for cues that may explain why some leaders seem more effective
than others. In the past 2 decades, scientists’ attention has been
turned to the effects of charismatic and transformational leader-
ship, pointing, among others, to the importance of the leader’s
focus on the collective (Bass, 1985; Bass & Avolio, 1993; Conger
& Kanungo, 1998; Shamir, House, & Arthur, 1993). In acknowl-
edgment of the notion that group-oriented leaders may be highly
effective, the present study focuses on the effects of self-
sacrificing behavior as a primary exemplar of such group-oriented
behavior (Choi & Mai-Dalton, 1998; Conger & Kanungo, 1987;
Shamir et al., 1993). This conceptualization of leader self-sacrifice
as group-oriented behavior invites an integration of these notions
from theories of charismatic and transformational leadership with
the social-identity analysis of leadership in which leader group-
orientedness is a core element (Hogg & van Knippenberg, 2003;
van Knippenberg & Hogg, 2003). On the basis of this social-
identity analysis, we argue that the characteristics of the leader as
a group member play a key role in leadership effectiveness and
that self-sacrifice may be especially important for leaders that are
not group prototypical (i.e., representative of the group; Hogg,
2001) of the group they supervise. In addition to pursuing this
integration of notions from theories of charismatic and transfor-
mational leadership with propositions from the social-identity
analysis, a secondary aim of the study is to provide first, and
causal, evidence that leader self-sacrifice may positively affect
follower performance.
Self-Sacrifice and Leadership Effectiveness
Many stories about great leaders, be it political leaders, military
leaders, religious leaders, or leaders of organizations, tell about the
self-sacrifice the leader was willing to make for the benefit of the
greater good. Clearly, people are inspired and astonished by self-
sacrificing behavior of others and assign meaning and purpose to
these acts. In general, it is assumed that these exceptional leaders
have a profound effect on their followers and, eventually, on social
systems.
Recently, leader self-sacrifice and the proposed effects of this
behavior have come under the increased attention of researchers
(Avolio & Locke, 2002; Choi & Mai-Dalton, 1998, 1999; de
Cremer, 2002; de Cremer & van Knippenberg, 2002; Yorges,
Weiss, & Strickland, 1999). Evidently, these researchers were
inspired by theories of charismatic and transformational leadership
(Bass, 1985; Bass & Avolio, 1993; Conger & Kanungo, 1987,
1998; House, 1977; Shamir et al., 1993). In general, transforma-
tional and charismatic leadership is considered to result in an
increased sense of collective identity and common mission, higher
leader effectiveness, greater levels of commitment and motivation,
the willingness to make personal sacrifices, and, ultimately,
heightened performance (Bass & Avolio, 1993; Lowe, Kroeck, &
Sivasubramaniam, 1996; Shamir et al., 1993). Although different
theories of charismatic and transformational leadership may have
their own explanation for why exactly these effects on followers
occur and may accentuate different aspects of leader behavior,
most of them emphasize behaviors of the leader that show his or
her dedication to the group and its mission (cf. Kark, Shamir, &
Chen, 2003). These leaders, for instance, show their group-
mindedness by making more references to the collective history,
the collective identity and interest, and collective efficacy than
Barbara van Knippenberg, Department of Work and Organizational
Psychology, Faculty of Psychology and Education, Vrije Universiteit,
Amsterdam, the Netherlands; Daan van Knippenberg, Rotterdam School of
Management, Erasmus University Rotterdam, Rotterdam, the Netherlands.
Netherlands Organization for Scientific Research (NWO) is gratefully
acknowledged for funding this project (MAGW, Grant 490 01–200). We
thank Steven Dumpel for his help with the data collection for Study 4.
Correspondence concerning this article should be addressed to Barbara
van Knippenberg, Faculty of Psychology and Education, Department of
Work and Organizational Psychology, van der Boechorststraat 1, 1081 BT
Amsterdam, the Netherlands. E-mail: bm.van.knippenberg@psy.vu.nl
Journal of Applied Psychology Copyright 2005 by the American Psychological Association
2005, Vol. 90, No. 1, 25–37 0021-9010/05/$12.00 DOI: 10.1037/0021-9010.90.1.25
25
noncharismatic and nontransformational leaders do (Shamir,
Arthur, & House, 1994; Shamir et al., 1993). Moreover, these
leaders have the ability to formulate a compelling vision concern-
ing the group’s ideal future, have the ability to shift group mem-
bers’ focus from self-interest to collective interest, and make
public demonstrations of their dedication to the cause (Conger,
1999; Lowe et al., 1996; Shamir et al., 1993). Jacobson and House
(2001) stated that “such demonstrations often involve significant
personal sacrifice or even danger, by which the leader projects an
image of courage, dedication, and commitment to the interests of
the collectivity” (p. 79). Indeed, being self-sacrificial is probably
one of the most direct ways for a leader to state that he or she
considers the group’s welfare to be important, and it also explicitly
shows his or her commitment to the collective. In sum, self-
sacrifice communicates the relatively unambiguous message that
the leader has a progroup orientation. It is not, therefore, surprising
that several researchers have suggested self-sacrifice as a typical
example of charismatic and transformational leadership behavior
(Bass, 1985; Choi & Mai-Dalton, 1998; Conger & Kanungo, 1987;
House & Shamir, 1993; Jacobson & House, 2001).
Accordingly, several authors have proposed that the extent to
which the leader engages in self-sacrificing behavior contributes to
positive follower perceptions (Bass, 1985; Choi & Mai-Dalton,
1999; Conger & Kanungo, 1987; de Cremer, 2002; Shamir et al.,
1993; Yorges et al., 1999). Choi and Mai-Dalton (1998) defined
leader self-sacrifice as the (total or partial) abandonment and/or
(permanent/temporary) postponement of personal interests, privi-
leges, or welfare in the division of labor, the distribution of
rewards, and/or the exercise of power. There are several well-
known anecdotes pertaining to the self-sacrificial behavior of
organizational leaders. For example, the CEO of a Dutch airline
company decided on a substantial cutback in his own salary when
faced with the airline company crisis after the attacks on Septem-
ber 11, 2001. Only then did he ask his employees to accept pay
restraint and a reduction of working hours. Other well known, but
more common, self-sacrificing behaviors include, for instance, a
supervisor’s willingness to take on a bigger part of the workload,
to forgo the right to a stylish and spacious office, or to give up a
day off in favor of a subordinate. The engagement in self-
sacrificing behavior is often seen as extraordinary and unconven-
tional behavior. As a consequence, leader self-sacrifice adds to
followers’ attribution of charisma to the leader.
Self-sacrifice seems to play an important part in group devel-
opment (Prapavessis & Carron, 1992) as well. As we argued
above, by making personal sacrifices, the leader clearly shows his
or her focus on the group’s welfare. Self-sacrifice thus not only has
short-term, direct positive consequences for the group’s function-
ing (i.e., the immediate benefits resulting from the self-sacrifice)
but also has a more long-term effect in the sense that it creates the
conviction among followers that the leader can be relied on to
behave in a group-oriented manner in future decisions as well (van
Lange et al., 1997). In this way, self-sacrifice may help build a
basis for leadership effectiveness that is more stable and enduring
and that goes beyond the situation in which the self-sacrifice was
made.
Moreover, and perhaps more important in light of a group’s
chances of survival when faced with competition or crisis, leader
self-sacrifice is also likely to lead to higher follower performance.
The reciprocity norm prescribes that people are under the pressure
to help those who have helped them. This norm may operate as a
behavioral rule that is present in ongoing relationships and that
helps in keeping them stable (Gouldner, 1960; Greenberg &
Folger, 1983). It may also operate as an intrinsic motive that
provides satisfaction by itself and that leads people to reciprocate
even when it is in conflict with immediate self-interest or when the
focal persons are not expecting to have an ongoing relationship
with each other (de Cremer & van Lange, 2001; Fehr & Schmidt,
2001; Gallucci & Perugini, 2003; Kahneman, 2003). A leader’s
self-sacrificing behavior will create the pressure on followers to do
as is done to them: forgo self-interest and commit oneself to the
collective. A self-sacrificing leader thus clearly prescribes what
kind of behavior is expected and appreciated in light of the group’s
common cause—or, as Mason Cooley once stated, “self-sacrifice
usually contains an unspoken demand for payment” (as cited in
Andrews, Biggs, & Seidel, 1996).
Although there is ample research on the relationship between
charisma on the one hand and subordinate perceptions of the leader
and leadership effectiveness on the other hand, most studies are
correlational in nature and do not explicitly focus on self-sacrifice
(Bass, 1990; Lowe et al., 1996; Shamir et al., 1993). The number
of experimental tests of the effects of charismatic and transforma-
tional leadership in general is much more limited (Barling, Weber,
& Kelloway, 1996; Dvir, Eden, Avolio, & Shamir, 2002; Howell
& Frost, 1989; Hunt, Boal, & Dodge, 1999; Kirkpatrick & Locke,
1996; Shea & Howell, 1999), especially the ones that focus di-
rectly on effects of leader self-sacrifice (Choi & Mai-Dalton, 1999;
de Cremer, 2002; de Cremer & van Knippenberg, 2002; Yorges et
al., 1999). These studies revealed that a self-sacrificing leader is
indeed perceived to be more charismatic than a self-benefiting
leader (Yorges et al., 1999) and that leader self-sacrifice elicits
stronger intentions to reciprocate the leader’s behavior (Choi &
Mai-Dalton, 1999), stronger feelings of group belongingness (de
Cremer & van Knippenberg, 2002), stronger intentions to contrib-
ute more money to a charity fund (Yorges et al., 1999), and more
cooperative behavior in a public-good dilemma (de Cremer, 2002;
de Cremer & van Knippenberg, 2002).
In short, the existing theoretical considerations suggest, and the
empirical evidence shows, that leader sacrifice renders followers
cooperative and motivated to reciprocate. In short, leader self-
sacrifice is associated with factors that determine leader effective-
ness. However, ultimately, when one wants to get down to the very
root of leadership effectiveness, one wants to know the effects of
leaders’ self-sacrifice on follower task performance. Yet, as far as
we know, there are no studies that focus on the question of whether
self-sacrificing leadership may indeed lead to higher follower
performance. Also given the importance of insight into perfor-
mance-enhancing factors for organizational practice, in the present
study, we aimed to test the hypothesis that self-sacrificing leaders
will be able to push subordinates to a higher performance level
than nonsacrificing leaders.
When exactly is self-sacrificial leadership successful and likely
to have the most impact? A partial answer to this question is rooted
in the analysis of self-sacrifice as communicating the leader’s
dedication to the group. If self-sacrifice reveals a group-
membership motive of the leader, its display would be especially
important under conditions that leave followers uncertain as to
whether or not the leader is committed to the group. Indeed, if
leader self-sacrifice is in part effective because it communicates to
26
VAN KNIPPENBERG AND VAN KNIPPENBERG
followers that the leader can be relied on to behave in a group-
oriented manner, its effects should be contingent on other factors
that may communicate leader group-orientedness (or, conversely,
that might raise doubts about leaders’ commitment to the group).
The social-identity analysis of leadership suggests that the extent
to which a leader is group prototypical, that is, representative of
the group’s identity (Hogg, 2001), is a core determinant of the
extent to which the leader is trusted to be group-oriented (Hogg &
van Knippenberg, 2003; van Knippenberg & Hogg, 2003). Ac-
cordingly, as we outline in the following sections, leader proto-
typicality may be an important moderator of the effectiveness of
leader self-sacrifice.
Leader Prototypicality and Leadership Effectiveness
Leader group prototypicality is seen as an important determinant
of leadership effectiveness in the social-identity analysis of lead-
ership (Hogg, 2001; Hogg & van Knippenberg, 2003; van Knip-
penberg & Hogg, 2003). This analysis has its origins in theories of
social identity and self-categorization (Ashforth & Mael, 1989;
Hogg & Abrams, 1988; Hogg & Terry, 2000; Tajfel & Turner,
1986; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). This
theoretical perspective describes how part of people’s self-
definition derives from their group memberships and is tied to
cognitive representations of these groups in the form of prototypes
(Rosch, 1978; Turner et al., 1987). These prototypes are context-
specific sets that define and prescribe beliefs, attitudes, norms,
values, and behavior (Hogg, 2001). The ingroup prototype is an
abstract cognitive representation of “us” that draws on intergroup
differences and ingroup similarity but also on group memory and
on past group history (van Knippenberg & Hogg, 2003). As a
function of the match between personal characteristics and group-
prototypical characteristics, some group members are more proto-
typical than others, just like some traits, attitudes, or behavioral
dispositions are more prototypical of the group than others. The
more prototypical the group member, the more he or she represents
the group’s standards, values, and norms. Prototypical group mem-
bers exemplify group normative behavior and reflect what mem-
bers of the ingroup have in common and what sets them apart from
relevant outgroups (Turner et al., 1987).
The social-identity analysis of leadership emphasizes that lead-
ers not only lead groups of people but also are members of these
groups: All organizational leaders are also a member of the orga-
nization, and of groups within the organization, and therefore share
one or more group memberships with the people they lead. Lead-
ership processes are thus enacted in the context of shared group
membership, and leaders’ characteristics as a group member may
therefore play an important role in leadership effectiveness. This
analysis suggests that a leader’s prototypicality of the group in
particular should be tied to leadership effectiveness because indi-
viduals that are more representative of the group are more influ-
ential and attractive (Hogg, 1992; van Knippenberg, Lossie, &
Wilke, 1994).
1
The proposition that leader prototypicality is a determinant of
leadership effectiveness is supported by an increasing number of
studies showing, for instance, that prototypical group members are
more likely to emerge as leaders (Fielding & Hogg, 1997; van
Knippenberg, van Knippenberg, & van Dijk, 2000) and that group-
prototypical leaders are more influential and effective (Hains,
Hogg, & Duck, 1997; Hogg, Hains, & Mason, 1998; Platow & van
Knippenberg, 2001; Platow, van Knippenberg, Haslam, van Knip-
penberg, & Spears, 2002; van Knippenberg & van Knippenberg,
2003). Evidence for the importance of leader prototypicality for
leader effectiveness and emergence was found in studies using
different paradigms, different operationalizations of prototypical-
ity, and different measures of leadership effectiveness. This sup-
port largely derives from experimental studies (Hains et al., 1997;
Hogg et al., 1998; Platow & van Knippenberg, 2001; Platow et al.,
2002) but also from a few studies in field settings (Fielding &
Hogg, 1997; van Knippenberg & van Knippenberg, 2003).
The notion that group prototypicality is a basis for leadership
effectiveness does by no means imply that prototypical leaders can
only behave like “the average group member.” The ingroup pro-
totype describes and prescribes group-membership appropriate
attributes in a specific context, with some attributes more central to
the group’s identity than others. The prototypical group member is
thus, in fact, closer to a representation of the ideal group member
than to the typical or average group member. It is therefore also
possible that the prototypical leader is, for instance, rather average
in some respects and exceptional or unconventional in others.
Prototypicality provides leaders with more leeway in their behav-
ior and thus positions them to effectively engage in behavior that
may lead the group or organization in new directions. Being
representative of the group’s identity as a basis of leadership
effectiveness is thus not at odds with engaging in special, unusual,
or distinctive behavior; indeed, it sets the stage for effectively
engaging in such behavior. In sum, displaying unusual and uncon-
ventional behavior, which is sometimes seen as a basis for lead-
ership effectiveness (Conger & Kanungo, 1987), is perfectly con-
sistent with maintaining (or even actively enhancing)
prototypicality.
Especially relevant to the present discussion is the fact that
prototypical group members are more likely to identify with the
group (i.e., the high degree of representativeness is likely to reflect
on an individual’s self-definition) and therefore should be more
likely to take the group’s interest to heart. As a result, leader
prototypicality and leader group-oriented behavior often go to-
gether and— especially important to the present analysis—will
also be expected to go together by subordinates. Whereas nonpro-
totypical leaders may not be considered to have the group’s best
interest at heart without concrete demonstrations to that end (i.e.,
the display of group-oriented behavior), prototypical leaders have
more leeway in their behavior because their prototypicality leads
their followers to have faith in the leader’s disposition to be
1
It is important to note, in order to avoid confusion, that the prototypi
-
cality concept also plays an important role in leadership-categorization
theory (e.g., Lord, Foti, & DeVader, 1984; Lord & Maher, 1991). Both
leadership-categorization theory and the social-identity approach base their
use of the term on the work of Rosch (1978), and, in that sense, the term
has identical meaning for both approaches. The important difference,
however, is that in leadership-categorization theory, leader prototypicality
refers to the leader’s representativeness of a leader category (i.e., matching
the stereotype of a leader), whereas in the social-identity analysis, leader
prototypicality reflects representativeness of the work group, team, or
organization that the leader is leading. Given that both uses of the concept
reflect a 20-year tradition (e.g., Lord et al., 1984; Turner, 1985), it seems
best to leave it at this double use of terminology.
27
LEADER SELF-SACRIFICE AND PROTOTYPICALITY
group-oriented (cf. van Knippenberg et al., 2000). Because the
group belongingness of less prototypical leaders is not so obvious,
nonprototypical leaders may actually need to engage in more
group-oriented behavior than prototypical leaders to be endorsed
as leaders (Platow & van Knippenberg, 2001). As a result, lead-
ership effectiveness should be more contingent on the extent to
which the leader engages in group-oriented behavior for less
prototypical leaders.
This analysis is congruent with classic research that shows that
legitimate and respected leaders are allowed a great deal of nor-
mative leeway in groups (e.g., Sherif & Sherif, 1964) and that
leaders who have earned credits in the eyes of their followers, for
instance by adherence to group norms, are allowed to behave
idiosyncratically and are allowed more freedom in performing
nonconformative actions (e.g., Hollander, 1958, 1992). The loyalty
of leaders who have not earned this credit is less evident; they
should therefore be more careful in displaying nonconformative
behavior and make certain to show their group-orientedness in
order to be endorsed and effective as leaders.
Present Study: Leader Self-Sacrifice and Leader
Prototypicality
As outlined in the above section, theories of charismatic and
transformational leadership suggest that leader self-sacrifice has
strong positive effects on leadership effectiveness in part because
it communicates that the leader is dedicated to the collective and
its goals and mission. The social-identity analysis shows that
because prototypical group leaders are, in a way, the embodiment
of the group, they are trusted to further the group’s interest without
having to display their group-orientedness (or at least to a lesser
extent than nonprototypical leaders). The situation is different for
less prototypical group leaders: Their group-orientedness is not
expected as a consequence of their representativeness, and by
making personal sacrifices for the group, they may enhance their
effectiveness substantially. On the basis, then, of the integration of
these notions from theories of charismatic and transformational
leadership and from the social-identity analysis of leadership, we
predict that leader prototypicality moderates the effects of leader
self-sacrifice on leadership effectiveness, such that the effects of
leader self-sacrifice are stronger for less prototypical leaders than
for more prototypical leaders. Accordingly, the following hypoth-
eses were tested in the present study:
Hypothesis 1: Self-sacrificing leaders are perceived as more
charismatic, more group-oriented, and more effective and
motivate higher follower performance then non-self-
sacrificing leaders.
Hypothesis 2: The effects of leader self-sacrifice are stronger
for less prototypical leaders than for more prototypical lead-
ers.
These hypotheses were tested first in a laboratory experiment
(Study 1). The advantage of this controlled experimental set-up is
that it yields results with high internal validity that make conclu-
sions concerning causality possible. Bringing the test of our hy-
potheses closer to leadership in actual organizations, but maintain-
ing the experimental nature of this test, we also conducted a
scenario experiment that allowed us to test our hypotheses exper-
imentally with more mundane realism (Study 2). To determine
whether the predicted relationships may also be observed in orga-
nizational settings, we also conducted two cross-sectional surveys
(Study 3 of business students working in a wide range of organi-
zations; Study 4 of employees of primary schools). In this way, the
strengths of the one method may compensate for the weaknesses of
the other (Dipboye, 1990).
Study 1
For Study 1, we adapted an experimental paradigm that has
established itself in over 30 years of research as the paradigm for
the experimental study of social categorization and social identity
(Brewer, 1979; Brewer & Brown, 1998; Tajfel, 1982), including
the role of leader prototypicality (Hains et al., 1997; Hogg et al.,
1998; Platow & van Knippenberg, 2001; van Knippenberg et al.,
2000) to study the effects of leader self-sacrifice and prototypical-
ity on follower performance and perceptions of leadership effec-
tiveness. In this paradigm, participants are allegedly assigned to a
group on the basis of a criterion that is either trivial or ambiguous
to ensure that groups do not exist before participating in the
experiment, and participants’ point of reference is formed by the
experimentally controlled information they receive about the
group and its members and not by prior experience with the group,
connotations of the categorization criterion, and so forth.
Participants in the present study were led by a (simulated) leader
who was either presented as prototypical or nonprototypical on the
basis of bogus feedback about test scores of the leader and the rest
of the group on a test unknown to participants (a test for brain-
hemispheric dominance; cf. Platow & van Knippenberg, 2001; van
Knippenberg & van Knippenberg, 2003; van Knippenberg et al.,
2000). To manipulate leader self-sacrifice, we suggested that the
leader did versus did not invest time and effort in the group task
above and beyond other duties assigned to the leader (cf. Choi &
Mai-Dalton, 1998). Participants were allegedly assigned to a group
to work on an idea-generation task in which the explicit production
goal was for each group member to produce as many ideas as
possible as contribution to the overall group production. Accord-
ingly, the number of ideas produced by the participant was our
measure of individual performance (cf. Diehl & Stroebe, 1987,
1991).
2
In addition, we assessed perceived leadership effective-
ness, leader group-orientedness, and leader charisma.
Method
Participants and Design
One-hundred seventy-four Dutch university students participated volun-
tarily in the study. Participants were randomly assigned to the conditions of
a 2 (self-sacrificing: yes vs. no) 2 (leader prototypicality: high vs. low)
between-subjects factorial design. Mean age was 20.71 years (SD 3.06),
28.2% of the participants were male, and 71.8% of the participants were
female.
2
Even though the explicit performance goal was quantity of ideas
generated, one may wonder about the quality of ideas generated. Previous
research using highly similar idea-generation tasks has shown that idea
quality and quantity are highly correlated (in the r .90 range; e.g., Diehl
& Stroebe, 1991). Therefore, we opted not to analyze the quality of ideas.
28
VAN KNIPPENBERG AND VAN KNIPPENBERG
Procedure
Participants were invited in small groups to take part in a study on
“group decision making and leader behavior.” Upon arrival in the labora-
tory, participants were seated in individual cubicles, each containing a
computer that was used to present all of the instructions, stimuli, and
questions and to register the dependent measures. Participants could not see
each other and believed that their reactions were anonymous. Strengthen-
ing the suggestion of interconnectedness and upcoming group decision-
making, we ensured that participants were asked to await the starting signal
so that all could enter the computer-mediated experiment simultaneously.
To manipulate leader prototypicality, we first asked participants to
individually complete a brain-hemisphere dominance test (adapted from
the Brain Works Test from Synergistic Learning Incorporated, 1997) and
then provided them with bogus feedback concerning their own and other
group members’ score on the hemispheric-dominance test. To make sure
that possible effects would not be caused by people believing left or right
hemispheric dominance is more positive (or more leader-like), we con-
trolled for possible effects of brain dominance feedback by counterbalanc-
ing over conditions whether participants learned that their group was
predominantly left- or right-hemisphere dominant. Bogus feedback about
group members’ test scores was given by representing each group member
by a letter from A–E and presenting the position of the five group members
on a scale ranging from left- to right-hemisphere dominant (participants
were always assigned the letter B and received a rather group-typical
score). A short clarification of the results was given, and some positive
characteristics of left or right brain hemisphere dominance were provided.
We then commenced with announcing the assignment of a group leader
who would be asked to supervise the upcoming group task.
Participants in the high-prototypical leader condition were told that we
opted to assign as leader the person who was most representative of the
group as a whole. We stressed the fact that the leader would be the person
who differed the least from the rest of the group members and who had
most in common with them. We then assigned Group Member E as the
group’s leader, whom feedback showed to be the most prototypical group
member. Participants in the low-prototypical leader condition were told
that we opted to assign as leader the person who was least representative
of the group as a whole. We now stressed the fact that the leader would be
the person who differed the most from and had least in common with the
other group members. We then assigned Group Member D as the group’s
leader, whom feedback showed to be the least prototypical group member
and the only one with deviating hemispheric dominance. In reality, Group
Members D and E did not exist, and the alleged leaders’ communications
consisted of preprogrammed messages.
As a second assignment, participants were asked to imagine that they
were part of a team that had to develop a campaign for the government. The
campaign was to heighten the awareness of possible individual contribu-
tions to environmental protection activities. Within the framework of the
campaign, participants were to partake in an idea-generation task. In 10
min of time, participants had to produce as many ideas as possible about
what they could do themselves to preserve the environment. Hence, the
required task was to generate as many ideas as possible. We told partici-
pants that the leader would get the opportunity to formulate specific
instructions, and a few moments later participants received the leader’s
message (note that this message was preprogrammed and that the leader
did not exist in reality).
In the self-sacrificing condition, participants read the following:
We have to come up with as many ideas as possible about what people
can do themselves for a better environment. We have ten minutes time
for this. I want to ask each of you to think of at least 30 ideas in those
10 minutes. I, as the leader, have gotten some other tasks as well. But
I find it important to participate in this task and set a good example.
I will therefore try to contribute at least 40 ideas to the group.
In the non-self-sacrificing conditions, participants received the follow-
ing message:
You have to come up with as many ideas as possible about what
people can do themselves for a better environment. You have ten
minutes time for this. I want to ask you to think of at least 30 ideas in
those 10 minutes. I, as the leader, have gotten some other tasks as
well. That’s why I find that I don’t have to perform this production
task. I will therefore not generate any ideas myself.
The participants then proceeded with the idea-generation task. The
number of ideas that each participant generated was the performance
measure that served as one of our measures of leadership effectiveness
(M 21.73, SD 8.04). After completing the group task, participants
were informed about our interest in their impressions so far and their
opinion about the leader specifically. Participants then filled out a ques-
tionnaire containing manipulation checks and additional measures. As a
check on our prototypicality manipulation, we asked participants to indi-
cate to what extent they considered their supervisor to be representative of
the group (1 absolutely not representative,7 absolutely very repre-
sentative) and to what extent they thought their supervisor resembled the
other group members (1 very little,7 very much). The correlation
between these two items (for similar measurements see Platow & van
Knippenberg, 2001) was high and significant (r .93, p .001). As a
check on our manipulation of self-sacrifice, we asked participants whether
or not their supervisor participated in the idea-generation task (1 yes,
2 no). To measure perceived leadership effectiveness, we asked partic-
ipants to indicate to what extent they agreed with four items, including “I
put my trust in this supervisor” and “My supervisor is an excellent
supervisor” (1 very much disagree,7 very much agree). Cronbach’s
alpha revealed good internal consistency for this scale (
.93, M 4.00,
SD 1.51). To measure perceived charisma of the leader, we formulated
questions inspired by Bass (1985) and Platow et al. (2002). This resulted in
a three-item scale (
.92, M 3.54, SD 1.76), with items such as
“This supervisor is enthusing,” and “This supervisor awakens my feelings
of commitment for the group” (1 very much disagree,7 very much
agree). The extent to which the leader was considered to be group oriented
was measured by participants’ reaction to three statements (
.95, M
4.13, SD 2.00), with items like “This leader exerts him/herself for the
benefit of the group” (1 not much at all, 7 very much so).
Results
In all analyses of variance (ANOVAs), self-sacrificing (yes/no),
leader prototypicality (high/low), and brain-hemispheric domi-
nance (left/right) were factors in the design.
Manipulation Checks
An ANOVA on participants’ perceptions of leader prototypical-
ity revealed a significant main effect for leader prototypicality,
F(1, 166) 252.39, p .0001,
2
.60, with the confidence
interval (CI) for the difference in observed unweighted means, or
CI(diff) between 3.17 and 2.47, indicating that those in the
high-prototypical leader condition perceived that the supervisor
was more prototypical than those in the low-prototypical leader
condition (Ms 5.15 vs. 2.33, respectively; SDs 1.05 vs. 1.26).
No other significant effects were found.
As a testament to the success of our self-sacrificing manipula-
tion, 95.4% of the participants picked the answer that corre-
sponded with the condition to which they were assigned (
.91,
p .0001).
29
LEADER SELF-SACRIFICE AND PROTOTYPICALITY
Dependent Measures
Our main dependent variables were actual performance (i.e., the
number of ideas generated) and the perceptions of leadership
effectiveness as measures of leadership effectiveness. In addition,
we assessed perceived charisma and leader group-orientedness.
We first performed a principal-components analysis with
OBLIMIN rotation of the items comprising the charisma, leader
group-orientedness, and perceived leadership effectiveness mea-
sures. This analysis yielded a three-factor solution, accounting for
88.6% of the variance, with all items loading above |.55| on the
intended component and all items loading at least |.20| lower on the
other components.
Leadership Effectiveness
Performance. An ANOVA on the number of ideas that were
generated revealed a main effect of self-sacrificing, F(1, 166)
11.80, p .001,
2
.07, CI(diff) between 6.35 and 1.71.
Participants who were led by a self-sacrificing leader generated
more ideas than participants who were led by a non-self-sacrificing
leader (Ms 23.76 vs. 19.65, SDs 8.11 vs. 7.45). As expected,
this effect was qualified by a Self-Sacrificing Leader Prototypi-
cality interaction, F(1, 166) 4.92, p .05,
2
.03, CI(diff)
between 4.92 and .28. Planned comparisons (Rosenthal &
Rosnow, 1985) showed that, in the low-prototypical leader condi-
tion, participants generated significantly more ideas when they had
a self-sacrificing leader than when they had a leader who was
non-self-sacrificing (M
non-self-sacrificing
18.61 vs. M
self-sacrificing
25.27, SDs 7.06 vs. 8.03), F(1, 172) 16.65, p .0001,
2
.09, CI(diff) between 9.96 and 3.42, whereas no such differ-
ence was found for participants in the high-prototypical group leader
condition (M
non-self-sacrificing
20.74 vs. M
self-sacrificing
22.19,
SDs 7.78 vs. 7.98), F(1, 172) .68, ns,
2
.00, CI(diff)
between 4.88 and 2.00.
Perceived leadership effectiveness. A main effect in the
ANOVA showed that the self-sacrificing leader was perceived to
perform better than the non-self-sacrificing leader (M
non-self-sacrificing
3.00 vs. M
self-sacrificing
4.97, SDs 1.21 vs. 1.08), F(1,
166) 128.16, p .0001,
2
.58, CI(diff) between 2.30
and 1.60. More important, we again found an interaction be-
tween self-sacrificing and leader prototypicality, F(1, 166) 5.16,
p .05,
2
.04, CI(diff) between .70 and .05, and the
pattern of results resembled the pattern of results for number of
ideas generated. When leaders were low prototypical, the partici-
pants regarded the self-sacrificing leader to perform better than
the non-self-sacrificing leader (M
non-self-sacrificing
2.74 vs.
M
self-sacrificing
5.10, SDs 1.20 vs. 1.06), F(1, 172) 77.91,
p .0001,
2
.31, CI(diff) between 2.88 and 1.83. How
-
ever, when the leader was high prototypical, performance evaluations
varied less with leader self-sacrifice (M
non-self-sacrificing
3.27 vs.
M
self-sacrificing
4.84, SDs 1.18 vs. 1.10), F(1, 172) 26.28, p
.0001,
2
.13, CI(diff) between 2.17 and 0.97.
Perceived Leader Charisma
Participants who were assigned to the self-sacrificing leader
condition indeed considered their supervisor to be more char-
ismatic than participants who were assigned to the non-self-
sacrificing leader condition (M
non-self-sacrificing
2.16 vs.
M
self-sacrificing
4.72, SDs 0.98 vs. 1.40), F(1, 166) 191.50,
p .0001,
2
.54, CI(diff) between 2.90 and 2.18.
Moreover, the expected interaction between self-sacrificing and
leader prototypicality emerged, F(1, 166) 4.09, p .05,
2
.03, CI(diff) between .73 and .01: Self-sacrificing leaders
were considered to be more charismatic than non-self-sacrificing
leaders in case of leader low prototypicality (M
non-self-sacrificing
2.07 vs. M
self-sacrificing
4.99, SDs 0.97 vs. 1.37), F(1, 172)
94.28, p .0001,
2
.35, CI(diff) between 3.51 and 2.33,
whereas this effect was smaller in the leader high prototypicality
condition (M
non-self-sacrificing
2.26 vs. M
self-sacrificing
4.43,
SDs 0.99 vs. 1.39), F(1, 172) 39.53, p .0001,
2
.18,
CI(diff) between 2.86 and 1.48.
Perceived Group-Orientedness
Participants who were assigned to the self-sacrificing leader
condition perceived the leader as more oriented toward group
benefit than participants who were assigned to the non-
self-sacrificing leader condition (M
non-self-sacrificing
2.38 vs.
M
self-sacrificing
5.85, SDs 1.03 vs. 0.95), F(1, 166) 548.63,
p .0001,
2
.77, CI(diff) between 3.76 and 3.18.
We also found an interaction between self-sacrificing and leader
prototypicality, F(1, 166) 8.54, p .005,
2
.05, CI(diff)
between .73 and .14, revealing that low-prototypical leaders
were considered to be more group oriented when they were
self-sacrificing than when they were not self-sacrificing
(M
non-self-sacrificing
2.10 vs. M
self-sacrificing
6.00, SDs 0.92
vs. 0.75), F(1, 172) 163.77, p .0001,
2
.49, CI(diff)
between 4.50 and 3.30, whereas this effect was less pronounced
in the high-prototypical leader condition (M
non-self-sacrificing
2.67
vs. M
self-sacrificing
5.70, SDs 1.03 vs. 1.10), F(1, 172) 67.43,
p .001,
2
.28, CI(diff) between 3.76 and 2.30.
Studies 2, 3, and 4
Study 1 showed that leader self-sacrifice had a positive effect on
performance—the first empirical demonstration of this effect—
and on perceptions of leadership effectiveness. It also showed that
leader self-sacrifice was related to perceptions of charisma and
group-orientedness. As expected, these effects were moderated by
leader prototypicality. The effects of self-sacrifice were stronger
for less prototypical leaders than for more prototypical leaders.
This supports our analysis of the effectiveness of leader self-
sacrifice and provides a first step in the integration of theories of
charismatic and transformational leadership with the social iden-
tity analysis of leadership.
Study 2, the scenario experiment, was designed to investigate
whether the effects of self-sacrifice and prototypicality on leader-
ship effectiveness may also be observed in an experimental setting
with more mundane realism. Studies 3 and 4, the correlational
surveys, allowed us to study people that actually participate in
organizations and to find out whether the hypothesized relation-
ships not only can occur but also do occur in reality (e.g., Bryman,
2000; Dipboye, 1990). Owing to important time constraints in the
administration of Studies 2 and 3, which were conducted in a
classroom context, we could only assess our main dependent
variable, (perceived) leadership effectiveness, in these studies.
30
VAN KNIPPENBERG AND VAN KNIPPENBERG
Study 3 had the advantage of sampling respondents from a wide
range of organizations but at the cost of control over the selection
of respondents (i.e., they self-selected by attending the lecture) and
having to rely on a student sample. Study 4, in contrast, drew its
sample from only one organization type but could rely on nonstu-
dent employees.
In Study 4, we were again able to assess leader charisma and
leader group-orientedness in addition to leadership effectiveness.
In addition, for another, less direct, measure of leader effective-
ness, we assessed employee’s willingness to participate in
organizational-change programs. The ability to engender change
and commit people to change is often seen as a key aspect of
effective leadership (Yukl, 2001).
Study 2
Method
Participants and Design
Four-hundred seventy-nine Dutch business students (67% male; mean
age 20.31 years, SD 1.25) participated voluntarily in the study as part
of a classroom demonstration. The design was a 2 (self-sacrificing: yes vs.
no) 2 (leader prototypicality: high vs. low) between-subjects factorial
design. Participants were randomly assigned to conditions (sample size per
analysis may differ because of missing data).
Procedure
Before participants listened to a lecture on organization theory, a short
business scenario was distributed. Participants were told that a situation in
which leadership plays a role would be portrayed, that they were to imagine
that they were in that particular situation, and that they were to answer the
questions accordingly. Participants read that they were to envision that
after they graduated, they went to work for an internationally oriented
consulting agency with a very good reputation. Participants were told that
this organization was widely seen as one of the best companies, if not the
best company, in its field and that there was a good match between them
and the company. They were to visualize that they worked in a team in
close collaboration with fellow team members: The coworkers were de-
scribed as people like themselves and as having the same attitude toward
life and work as the participants had.
In the low-prototypical leader condition, participants were informed that
the team leader was somewhat of an “outsider” in the team, that he was
very different from other team members and that he or she had a different
background, different interests, and a different attitude toward life and
work than the other team members. In the high-prototypical leader con-
dition, participants were informed that the team leader was very represen-
tative of the kind of persons in the team, that he was very similar to the
other team members, and that he had a similar background, similar inter-
ests, and a similar attitude toward life and work as the other team members.
Hereafter, participants were informed that as with all companies in industry
these days, the company where they worked was faced with the conse-
quences of the economic setbacks being experienced worldwide. To com-
bat the consequences of these setbacks, the company’s management de-
cided to cut salary costs. It was up to each team leader to decide how to
make these cuts for their team, as long as the total salary costs for each
team decreased. The following paragraph described the leader’s behavior
as either self-sacrificial or non-self-sacrificial. As in Choi and Mai-Dalton
(1999), self-sacrificial leader behavior was manipulated by adding inci-
dents of self-sacrificial behavior to the behavior prescribed for the non-
self-sacrificial leader. In the non-self-sacrifice condition, participants were
told: “Your team leader informs your team of the need to cut salary costs.
He asks all team members to accept a salary cut of 10%.” In the self-
sacrifice condition, the following sentence was added: “To protect the
salary of the team members as much as possible, he has cut his own salary
by 20%.”
Dependent Measures
All responses were assessed on 5-point disagree–agree scales. The leader
prototypicality manipulation was checked with three items (
.94; M
2.80, SD 1.51), “This team leader is a good example of the kind of
people that are members of my team,” “This team leader represents what
is characteristic about the team,” and “This team leader has a lot in
common with the members of the team” (cf. Platow & van Knippenberg,
2001; van Knippenberg & van Knippenberg, 2003).
The leader self-sacrifice manipulation was checked with two items,
“This team leader makes a personal sacrifice for the team,” and “This team
leader sacrifices salary in the team’s interest” (
.70; M 3.23, SD
1.20).
Leadership effectiveness was assessed with seven items, including “This
team leader is effective as a leader” and “This team leader is a good team
leader” (
.91; M 3.13, SD 0.85).
Results
Manipulation Checks
An ANOVA on the check on the prototypicality manipulation
indicated that the prototypical leader was rated as more prototyp-
ical (M 4.19, SD 0.71) than the nonprototypical leader (M
1.48, SD 0.64), F(1, 468) 1910.25, p .0001,
2
.80,
CI(diff) between 2.82 and 2.58. No other effects were
significant.
An ANOVA on the check on the self-sacrifice manipulation
showed that the self-sacrificing leader was seen as more self-
sacrificing (M 4.06, SD 0.82) than the nonsacrificing leader
(M 2.39, SD 0.89), F(1, 468) 448.90, p .0001,
2
.49,
CI(diff) between 1.82 and 1.52. No other effects emerged.
We may thus conclude that our manipulations were successful.
Perceived Leadership Effectiveness
An ANOVA on the perceived leadership effectiveness scale
yielded both significant main effects of leader self-sacrifice, F(1,
474) 5.35, p .025,
2
.01, CI(diff) between .33, and
.03, and leader prototypicality, F(1, 474) 8.53, p .005,
2
.02, CI(diff) between .38 and .07, and a significant Sacri-
fice Prototypicality interaction, F(1, 474) 4.35, p .05,
2
.01, CI(diff) between .31 and .01. Self-sacrificing leaders
were seen as more effective (M 3.22, SD 0.88) than nonsac-
rificing leaders (M 3.04, SD 0.82), and prototypical leaders
were seen as more effective (M 3.25, SD 0.85) than nonpro-
totypical leaders (M 3.02, SD 0.84). Of primary importance,
the effect of self-sacrifice was only significant in the nonproto-
typical leader condition (M 3.19 vs. M 2.85, SDs 0.84 vs.
0.82), F(1, 474) 9.97, p .005,
2
.02, CI(diff) between
.55 and .13, and not in the prototypical leader condition (M
3.25 vs. M 3.24, SDs 0.92 vs. 0.77), F 1.
31
LEADER SELF-SACRIFICE AND PROTOTYPICALITY
Study 3
Method
Sample
Respondents were 193 business students in a course on organizational
behavior, who participated voluntarily as part of a classroom demonstra-
tion. At the beginning of the course, a questionnaire assessing the study
variables was administered. As a consequence of the nature of the sample,
respondents were from a wide range of organizations. Sixty-seven percent
of the respondents were male, mean age was 22 (SD 2.16), and average
tenure on the job was 1.6 years (SD 1.74). All participants were
employed and reported about their direct supervisor in this job.
Independent Measures
All responses were assessed on 5-point scales (1 disagree,5 agree).
Leader self-sacrifice. Perceived leader self-sacrifice was assessed with
five items inspired by the work of Conger and Kanungo (1998) and Choi
and Mai-Dalton (1998). Items included “My supervisor is willing to stand
up for the team’s interest, even at the expense of his/her own interest” and
“My supervisor is always among the first to sacrifice free time, privileges,
or comfort if that is important for the team’s mission.”
Leader prototypicality. Prototypicality was assessed with five items
based on the works of Platow and van Knippenberg (2001) and van
Knippenberg and van Knippenberg (2003). The scale included the three
items that were used in Study 2 and had two additional items.
Dependent Measure
Perceived leadership effectiveness. Effectiveness perceptions were as-
sessed with three items similar to those used in the other studies. An
example of one of the items is “My supervisor is a good supervisor.”
Results
We first performed a principal-components analysis with OB-
LIMIN rotation of the items comprising our independent variables
(i.e., self-sacrifice and prototypicality). This analysis yielded a
two-factor solution, accounting for 70% of the variance, with all
items loading |.70| or higher on the intended scale and all cross-
loadings below |.30|. Means, standard deviations, and intercorre-
lations for the study variables are displayed in Table 1.
To test our hypotheses, we conducted a hierarchical regression
analysis. Following Aiken and West (1991), self-sacrifice and
prototypicality were centered (i.e., by subtracting the mean from
each score), and the interaction term as well as the main effects
were based on these centered scores. In the first step, we entered
leader self-sacrifice and prototypicality, as well as tenure, which
was added as a nuisance variable. In Step 2, the interaction term
was entered. Table 2 shows the regression results. After Step 2, the
main effects of self-sacrifice and prototypicality and the interac-
tion were significant. As expected, both leader self-sacrifice and
leader prototypicality were positively related to leadership effec-
tiveness. To further analyze the interaction, we conducted simple
slopes analysis (Aiken & West, 1991). Although self-sacrifice was
significantly related to leadership effectiveness both when leader
prototypicality was high (1 SD above the mean),
.40, p
.001, and when leader prototypicality was low (1 SD below the
mean),
.57, p .001, the relationship was more pronounced
when the leader was relatively low in prototypicality.
Study 4
Method
Sample
Respondents were 161 employees of primary schools belonging to the
IJmond-North primary school district of the Netherlands. Questionnaires
were sent to 512 potential respondents (response rate 31.3%), with the
instruction to either mail the completed questionnaire to the handling
researcher or to leave the questionnaire at the administration desk. Sixteen
percent of the respondents were male, and the mean age was 41 (SD
10.47).
Independent Measures
All responses were assessed on 5-point scales (1 disagree,5 agree).
Leader self-sacrifice. Perceived leader self-sacrifice was assessed with
the same items that were used in Study 3.
Leader prototypicality. Prototypicality was assessed with six items.
The items that were used in Study 3 were supplemented with an extra item
in order to explicitly include the aspect that the prototypical leader exem-
plifies group normative behavior in the prototypicality scale. The extra
added item was “This supervisor is an embodiment of our group norms.”
Dependent Measures
Perceived leadership effectiveness. Effectiveness perceptions were as-
sessed with four items, similar to the items used in the previous studies.
Willingness to engage in organizational change. For a less direct
measure of leadership effectiveness, we assessed employees’ willingness to
participate in organizational-change programs that were to be implemented
within the school district. The items were “I support plans that aim to
develop this organization” and “I am willing to cooperate completely in
order to implement change plans.”
Perceived leader charisma. Perceived leader charisma was assessed
with the same items that were used in Study 1, plus an extra added item.
This item was “This supervisor is inspiring.”
Perceived group-orientedness. Leader group-orientedness was as-
sessed with four items, similar to the items used in our laboratory study. An
example of an item is “My supervisor focuses first and foremost on the
team’s interest.”
Results
We first performed a principal-components analysis with
OBLIMIN rotation of the items of the predictor variables (i.e.,
self-sacrifice and prototypicality measures). This analysis yielded
a two-factor solution, accounting for 71% of the variance, with all
Table 1
Means, Standard Deviations, and Intercorrelations for Study 3
Variable MSD 1234
1. Self-sacrifice 3.14 0.89 (.85)
2. Prototypicality 3.03 0.99 .51*** (.92)
3. Tenure 1.60 1.75 .13 .03
4. Leadership effectiveness 3.23 1.02 .70*** .64*** .02 (.83)
Note. Coefficients alpha are displayed on the diagonal. N 185 (list-
wise).
*** p .001.
32
VAN KNIPPENBERG AND VAN KNIPPENBERG
items loading above |.73| on the intended scale and all crossload-
ings below |.10|. A second principal-components analysis of the
items of the dependent variables (i.e., effectiveness, willingness to
engage in organizational change, charisma, and group-
orientedness) yielded a four-factor solution, accounting for 81% of
the variance, with all items loading above |.64| on the intended
scale and all crossloadings below |.35|. Means, standard devia-
tions, and intercorrelations for the study variables are displayed in
Table 3.
To test our hypotheses, we conducted hierarchical regression
analyses. Following Aiken and West (1991), self-sacrifice and
prototypicality were centered (i.e., by subtracting the mean from
each score), and the interaction term as well as the main effects
were based on these centered scores. Table 4 shows the regression
results for all the dependent variables. When perceived leadership
effectiveness was the dependent variable, the main effects of
self-sacrifice and prototypicality, as well as the interaction, were
significant. As expected, both leader self-sacrifice and leader pro-
totypicality were positively related to leadership effectiveness. To
further analyze the interaction, we conducted simple slopes anal-
yses (Aiken & West, 1991). The relationship between self-
sacrifice and leadership effectiveness was more pronounced when
the leader was relatively low in prototypicality (1 SD below the
mean),
.66, p .001, as compared with when the leader was
high in prototypicality (1 SD above the mean),
.40, p .001.
When willingness to engage in organizational change was the
dependent variable, the results revealed a main effect of self-
sacrifice, showing that leader self-sacrifice was positively related
to leadership effectiveness. Also, we again found that the interac-
tion was significant. Simple slopes analyses revealed that the
relationship between self-sacrifice and willingness to change was
significant when the leader was relatively low in prototypicality (1
SD below the mean),
.31, p .005, and not significant when
the leader was high in prototypicality (1 SD above the mean),
.18, ns.
Another regression analysis was conducted with leader charisma
as the dependent variable. The main effects of self-sacrifice and
prototypicality, as well as the interaction, were significant. Both
leader self-sacrifice and leader prototypicality were positively re-
lated to leader charisma. The simple slopes analyses revealed again
that the relationship between self-sacrifice and leader charisma
was stronger when the leader was relatively low in prototypicality
(1 SD below the mean),
.53, p .001, as compared with when
the leader was high in prototypicality (1 SD above the mean),
.21, p .05.
Last, we performed a regression analysis with perceived leader
group-orientedness as the dependent variable. Again, leader self-
sacrifice and leader prototypicality were positively related to
leader group-orientedness, and the simple slopes analyses con-
ducted to further analyze the interaction revealed that the relation-
ship between self-sacrifice and leader group-orientedness was sig-
nificant when the leader was relatively low in prototypicality (1 SD
below the mean),
.45, p .001, and not significant when
the leader was high in prototypicality (1 SD above the mean),
.14, ns.
General Discussion
Leader self-sacrifice has been suggested to be a particularly
effective act of leadership, at least in part because it communicates
the leader’s dedication to the collective. We proposed that the
leader’s prototypicality of the collective moderates the effects of
leader self-sacrifice on leadership effectiveness because leader
prototypicality raises trust in leader’s group-orientedness and
should therefore render leadership endorsement and effectiveness
less contingent on the display of group-oriented behavior like
leader self-sacrifice. These predictions, derived from the integra-
tion of analyses of charismatic and transformational leadership
(Bass, 1985; Bass & Avolio, 1993; Conger & Kanungo, 1998;
Shamir et al., 1993) with the social-identity analysis of leadership
(Hogg & van Knippenberg, 2003; van Knippenberg & Hogg,
2003), were put to the test in a series of four studies that yielded
Table 2
Results of Hierarchical Regression Analysis for Study 3
Predictor bSEb
CI
R
2
tLower Upper
Self-sacrifice .56 .06 .43 .68 .49 .30 8.84**
Prototypicality .39 .06 .28 .50 .38 .11 7.06**
Tenure .01 .02 .05 .04 .02 .00 0.31
Self-Sacrifice
Prototypicality .10 .05 .19 .00 .10 .03 2.02*
Note. N 185 (listwise). Entries are statistics for Step 2. CI lower and CI
upper represent the 95% confidence intervals for the unstandardized re-
gression coefficients. R
2
is the variance explained by each predictor after
the other predictors have been entered into the equation.
* p .05. ** p .01.
Table 3
Means, Standard Deviations, and Intercorrelations for Study 4
Variable MSD 123456
1. Self-sacrifice 3.71 .84 (.87)
2. Prototypicality 3.47 .89 .50*** (.93)
3. Leader effectiveness 4.07 .80 .48*** .54*** (.91)
4. Willingness to change 4.41 .67 .34*** .23 .43*** (.88)
5. Leader charisma 3.82 .95 .59*** .60*** .73*** .42*** (.91)
6. Group-orientedness 4.26 .75 .62*** .54*** .65*** .46*** .59*** (.90)
Note. Coefficients alpha are displayed on the diagonal. N 154 (listwise).
*** p .001.
33
LEADER SELF-SACRIFICE AND PROTOTYPICALITY
consistent evidence in support of our hypotheses. The four studies
showed that the effects of leader self-sacrificing behavior on
leadership effectiveness were stronger for leaders who were less
prototypical of the group than for leaders who were more proto-
typical. Confidence in our findings is bolstered not just by the
replication over studies per se but especially by the fact that the
studies used different methodologies (i.e., laboratory experiment,
scenario experiment, cross-sectional survey), different samples
(i.e., students, nonstudent employees of organizations), and differ-
ent operationalizations of leadership effectiveness (i.e., follower
performance, perceived effectiveness, willingness to engage in
organizational change).
Further corroborating our analysis, Studies 1 and 4 showed that
leader self-sacrifice and prototypicality interacted in their relation-
ship with follower perceptions of leader group-orientedness. As
predicted, leader self-sacrifice had a positive effect on perceptions
of leader group-orientedness, and this effect was stronger for less
prototypical leaders. An interesting implication of our theoretical
analysis and these empirical findings is that other aspects of
leadership that affect perceptions of leader group-orientedness
may similarly moderate the effectiveness of leader self-sacrifice.
The effects of self-sacrifice might, for instance, also be contingent
on (a) whether or not the leader favors his or her own group over
other groups in allocation decisions (Platow & van Knippenberg,
2001), (b) a leader’s expressions of commitment to the group (de
Cremer & van Vugt, 2002), and (c) a leader’s appeals to the
group’s interest rather than individual self-interest (Platow et al.,
2002).
In addition, Studies 1 and 4 show that leader prototypicality
moderates the effect of leader self-sacrifice on perceptions of
charisma. This finding adds support to the proposition that leader
self-sacrifice leads to attributions of charisma (Choi & Mai-
Dalton, 1998, 1999; Conger & Kanungo, 1987; Shamir et al.,
1993; Yorges et al., 1999) and suggests that the integration of
propositions from theories of charismatic and transformational
leadership with propositions from the social-identity analysis of
leadership may not just advance our understanding of leadership
effectiveness per se but also our understanding of charismatic and
transformational leadership (cf. van Knippenberg & Hogg, 2003).
Although the main focus of the present study was on the
interactive effect of leader self-sacrifice and leader prototypicality,
the evidence obtained for the main effects of self-sacrifice and
prototypicality also is of interest. All four studies supported the
prediction that leader self-sacrifice contributes to leadership effec-
tiveness. Leader self-sacrifice not only resulted in the perception of
leadership effectiveness and the willingness to engage in organi-
zational change, it also positively affected follower performance
(i.e., Study 1). Whereas the effects on leadership perceptions may
be seen as similar to findings in earlier studies (Choi & Mai-
Dalton, 1999; de Cremer, 2002; de Cremer & van Knippenberg,
2002; Yorges et al., 1999), Study 1 extends these earlier findings
by yielding the first evidence for the performance-enhancing effect
of leader self-sacrifice. This finding is important because percep-
tual measures of leadership effectiveness cannot be equated with
the actual behavioral effects of leadership (Lord & Maher, 1991),
and the behavioral effects of leadership are typically more of a
concern to organizations than the perceptual effects. Study 1 thus
yields important new evidence for the presumed effectiveness of
leader self-sacrifice.
Most experimental studies of leader self-sacrifice have yielded
evidence of the effectiveness of leader self-sacrifice in comparison
with leader self-benefiting behavior (de Cremer, 2002; de Cremer
& van Knippenberg, 2002; Yorges et al., 1999). This raises the
question, however, of whether self-sacrifice has positive effects, or
self-benefiting has negative effects, or both. In other words, these
findings offer no basis for the conclusion that self-sacrifice per se
enhances leadership effectiveness. For instance, Yorges et al.
(1999) expected to find a positive effect of leader self-sacrifice on
charisma ratings when contrasted with a control group, but they
failed to find a significant difference between these two conditions
and only found a significant negative effect of self-benefiting on
charisma. They argued that they expected to find a more symmet-
rical effect in any replication, such that self-sacrificing behavior
would lead to higher charisma ratings. The present experiments
Table 4
Results of Hierarchical Regression Analyses for Study 4
Dependent variable Predictor bSEb
CI
R
2
tLower Upper
Leader effectiveness Self-sacrifice .16 .07 .03 .30 .17 .04 2.35**
Prototypicality .48 .06 .36 .61 .53 .28 7.56**
Self-Sacrifice Prototypicality .14 .06 .25 .03 .16 .04 2.57**
Willingness to change Self-sacrifice .16 .06 .03 .30 .20 .04 2.33*
Prototypicality .04 .06 .08 .18 .07 .00 .77
Self-Sacrifice Prototypicality .21 .06 .33 .10 .30 .09 3.80**
Leader charisma Self-sacrifice .38 .08 .22 .54 .33 .13 4.64**
Prototypicality .40 .08 .25 .56 .37 .16 5.21**
Self-Sacrifice Prototypicality .20 .07 .33 .07 .19 .06 3.05**
Leader group-orientedness Self-sacrifice .36 .06 .24 .49 .29 .18 5.72**
Prototypicality .25 .06 .13 .36 .29 .11 4.26**
Self-Sacrifice Prototypicality .15 .05 .26 .05 .18 .06 2.99**
Note. Ns between 147 and 149. CI lower and CI upper represent the 95% confidence intervals for the unstandardized regression coefficients. R
2
is the
variance explained by each predictor after the other predictors have been entered into the equation.
* p .05. ** p .01.
34
VAN KNIPPENBERG AND VAN KNIPPENBERG
(i.e., Studies 1 and 2) contrasted a leader self-sacrifice condition
with a no-self-sacrifice rather than a self-benefiting condition and
showed that, indeed, self-sacrifice can have a positive effect on
leadership effectiveness and perceptions of charisma. In this sense
too, the present study yields new evidence for the proposed effec-
tiveness of leader self-sacrifice.
We should be careful, however, to conclude that being self-
sacrificial is a sure route to leadership effectiveness and positive
follower perceptions. There may be a point where an increase in
leader sacrifice may have little extra positive impact on followers
and the organization but at the same time may have large negative
consequences for the leader him- or herself. Evidently, then, we do
not mean to imply that leaders must be extremely self-sacrificial or
must transcend their self-interest every day. Rather, we suggest
that leaders must be able and willing to display self-sacrificing
behavior from time to time (Avolio & Locke, 2002).
Three of the four studies revealed a main effect of prototypi-
cality on leadership effectiveness ratings, and Study 4 also yielded
main effects of prototypicality on charisma and group-
orientedness. More prototypical leaders were perceived to be more
effective, more charismatic, and more group-oriented than less
prototypical leaders. These effects are in line with predictions from
the social-identity analysis of leadership. Although there already
was substantial evidence for the effects of leader prototypicality,
most of this evidence derived from laboratory experiments or other
studies with nonorganizational samples (Hogg & van Knippen-
berg, 2003; van Knippenberg & Hogg, 2003). The present study
thus provides an important replication of these findings in organi-
zational settings.
Group prototypes, and therefore also the relative prototypicality
of members, are not fixed. What or who is prototypical is depen-
dent on intergroup comparisons (i.e., because the prototype is in
part defined by what distinguishes the group from relevant com-
parison groups; Rosch, 1978; Turner et al., 1987). Therefore, as the
comparative context changes, for example when comparisons are
made with a different outgroup, the prototype may change and,
accordingly, the relative prototypicality of specific group members
may change. In addition, group members may actively display or
manipulate their own prototypicality, for instance, by actively
seeking comparison with other outgroups or by simply stating that
they represent the group’s identity (Reicher & Hopkins, 2001,
2003). Because the extent to which a person is prototypical is a
dynamic feature, it is also possible that, under the right circum-
stances, the display of self-sacrificial behavior influences leader
prototypicality. For instance, in groups where self-sacrifice is the
norm (e.g., the Salvation Army, religious groups like the brothers
of the Franciscan order, various military units, or terrorist groups),
the display of self-sacrificial or unselfish behavior may enhance
the person’s representativeness. However, in most groups, self-
sacrifice does not belong to the core values of the group,
making self-sacrifice something that not necessarily enhances
prototypicality.
Of course, the present study is not without its weaknesses and
limitations. Three out of four studies relied on student samples.
This should not be considered problematic for experimental stud-
ies that are aimed at establishing causality in relationships with
high internal validity, and there is no reason to expect students to
behave differently from other populations (Brown & Lord, 1999;
Dipboye, 1990; Wofford, 1999). Even so, it should be noted that
Study 4, the nonstudent sample, did not include a performance
measure, so we could not completely replicate Study 1. An im-
portant direction for future research would therefore be to test the
interaction between leader self-sacrifice and prototypicality in a
field study with actual performance measures. A study along those
lines would also be able to address limitations associated with the
fact that Studies 3 and 4 had a mono-source and mono-method
design and relied on leadership perceptions. Such a design is
associated with two problems. First, it may inflate relationships
between variables. The main effects of self-sacrifice and proto-
typicality observed in the surveys may therefore be overestimated.
Therefore, even though we also obtained experimental evidence
for these main effects, it would be valuable if a future study in the
field tested these relationships with a design that does not suffer
from this problem. It is important to note, however, that common
source or method bias cannot account for statistical interactions—
indeed, because it may inflate main effects, it leads to an under-
estimation of the effect size of interactions and lowers the power
for the test of interactions (Evans, 1985; McClelland & Judd,
1993)—and thus forms no threat to the validity of our conclusions
about the Leader Self-Sacrifice Prototypicality interaction. Sec-
ond, common method bias set aside, perceptions of leadership
effectiveness may be affected by other influences than objective
indications of leadership effectiveness alone (Lord & Maher,
1991). For this reason too, then, replication of the performance
findings of Study 1 in an organizational context would bolster our
confidence in the generalizability of our findings.
It may also be noted that Study 1, although in one sense the
strongest study given the experimental evidence for performance
effects, is in another sense the weakest with its smallest mundane
realism. At the same time, however, we may note that not only
were the results of Study 1 replicated in Studies 2, 3, and 4, but
also that leadership experiments using paradigms similar to that of
Study 1 have consistently yielded results that generalize to sce-
nario and field settings (de Cremer & van Knippenberg, 2002; de
Cremer, van Knippenberg, van Knippenberg, Mullenders, &
Stinglhamber, 2005; van Knippenberg & van Knippenberg, 2003).
This thus suggests a greater mundane value than may appear at
first sight.
Because leader self-sacrifice may be regarded as exemplary of
the group-oriented aspects of charismatic and transformational
leadership, the current findings for the effects of self-sacrifice and
the role of leader prototypicality as moderator of the effectiveness
of self-sacrifice may advance our understanding of charismatic
leadership and leadership effectiveness. This is not to say, how-
ever, that leader prototypicality, and other factors that may affect
perceived leader group-orientedness, are the only moderators or
necessarily the most important moderators of the effectiveness of
leader self-sacrifice (cf. de Cremer, 2002). The role of leader
prototypicality set aside, there may be important boundary condi-
tions to the effectiveness of leader self-sacrifice, and a more
complete assessment of the role and importance of leader self-
sacrifice would therefore require studies that focus on a broader
range of leadership aspects and take a range of situational and task
contingencies into account. To yield a more complete picture of
the effectiveness of leader self-sacrifice and other group-oriented
aspects of charismatic leadership, it would therefore seem highly
worthwhile to pursue the study of these potential contingencies in
future research.
35
LEADER SELF-SACRIFICE AND PROTOTYPICALITY
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting interactions. New York: Sage.
Andrews, R., Biggs, M., & Seidel, M. (1996). The Columbia world of
quotations. New York: Columbia University Press.
Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organi-
zation. Academy of Management Review, 14, 20–39.
Avolio, B., & Locke, E. A. (2002). Contrasting different philosophies of
leader motivation: Altruism versus egoism. The Leadership Quarterly,
13, 169 –191.
Barling, J., Weber, T., & Kelloway, E. K. (1996). Effects of transforma-
tional leadership training on attitudinal and financial outcomes: A field
experiment. Journal of Applied Psychology, 81, 827–832.
Bass, B. M. (1985). Leadership and performance beyond expectations.
New York: Free Press.
Bass, B. M. (1990). Bass & Stogdill’s handbook of leadership (3rd ed.).
New York: Free Press.
Bass, B. M., & Avolio, B. (1993). Transformational leadership: A response
to critiques. In M. M. Chemers & R. Ayman (Eds.), Leadership theories
and research: Perspectives and directions (pp. 49 80). New York:
Academic Press.
Brewer, M. B. (1979). In-group bias in the minimal group situation: A
cognitive-motivational analysis. Psychological Bulletin, 86, 307–324.
Brewer, M. B., & Brown, R. J. (1998). Intergroup relations. In D. T.
Gilbert. S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychol-
ogy (4th ed., pp. 554–594). Boston: McGraw Hill.
Brown, D. J., & Lord, R. G. (1999). The utility of experimental research in
the study of transformational/charismatic leadership. The Leadership
Quarterly, 10, 531–539.
Bryman, A. (2000). Research methods and organization studies. London:
Routledge.
Choi, Y., & Mai-Dalton, R. R. (1998). On the leadership function of
self-sacrifice. The Leadership Quarterly, 9, 475–501.
Choi, Y., & Mai-Dalton, R. R. (1999). The model of followers’ responses
to self-sacrificial leadership: An empirical test. The Leadership Quar-
terly, 10, 397– 421.
Conger, J. A. (1999). Charismatic and transformational leadership in
organizations: An insider’s perspective on these developing streams of
research. The Leadership Quarterly, 10, 145–179.
Conger, J. A., & Kanungo, R. N. (1987). Toward a behavioral theory of
charismatic leadership in organizational settings. Academy of Manage-
ment Review, 12, 637– 647.
Conger, J. A., & Kanungo, R. N. (1998). Charismatic leadership in
organizations. Thousand Oaks, CA: Sage.
de Cremer, D. (2002). Charismatic leadership and cooperation in social
dilemmas: A matter of transforming motives? Journal of Applied Social
Psychology, 32, 997–1016.
de Cremer, D., van Knippenberg, B., van Knippenberg, D., Mullenders, D.,
& Stinglhamber, F. (2005). Rewarding leadership and fair procedures as
determinants of self-esteem. Journal of Applied Psychology, 90, 3–12.
de Cremer, D., & van Knippenberg, D. (2002). How do leaders promote
cooperation? The effects of charisma and procedural fairness. Journal of
Applied Psychology, 87, 858 866.
de Cremer, D., & van Lange, P. A. M. (2001). Why prosocials exhibit
greater cooperation than proselfs: The roles of social responsibility and
reciprocity. European Journal of Social Psychology, 15, 5–18.
de Cremer, D., & van Vugt, M. (2002). Intergroup and intragroup aspects
of leadership in social dilemmas: A relational model of cooperation.
Journal of Experimental Social Psychology, 38, 126–136.
Diehl, M., & Stroebe, W. (1987). Productivity loss in brainstorming
groups: Toward the solution of a riddle. Journal of Personality and
Social Psychology, 53, 497–509.
Diehl, M., & Stroebe, W. (1991). Productivity loss in idea-generating
groups: Tracking down the blocking effect. Journal of Personality and
Social Psychology, 61, 392– 403.
Dipboye, R. L. (1990). Laboratory vs. field research in industrial and
organizational psychology. In C. L. Cooper & I. T. Robertson (Eds.),
International review of industrial and organizational psychology (Vol.
5, pp. 1–34). Chichester, United Kingdom: Wiley.
Dvir, T., Eden, D., Avolio, B. J., & Shamir, B. (2002). Impact of trans-
formational leadership on follower development and performance: A
field experiment. Academy of Management Journal, 45, 735–744.
Evans, M. G. (1985). A Monte Carlo study of the effects of correlated
method variance in moderated regression analysis. Organizational Be-
havior and Human Decision Processes, 36, 305–323.
Fehr, E., & Schmidt, K. M. (2001). Theories of fairness and reciprocity:
Evidence and economic applications. London: Centre for Economic
Policy Research.
Fielding, K. S., & Hogg, M. A. (1997). Social identity, self-categorization,
and leadership: A field study of small interactive groups. Group Dy-
namics: Theory, Research, and Practice, 1, 39–51.
Gallucci, M., & Perugini, M. (2003). Information seeking and reciprocity:
A transformational analysis. European Journal of Social Psychology, 33,
473– 495.
Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement.
American Sociological Review, 25, 161–178.
Greenberg, J., & Folger, R. (1983). Procedural justice, participation, and
the fair process effect in groups and organizations. In P. B. Paulus (Ed.),
Basic group processes (pp. 235–256). New York: Springer-Verlag.
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, 1087–1100.
Hogg, M. A. (1992). The social psychology of group cohesiveness: From
attraction to social identity. London: Harvester Wheatsheaf.
Hogg, M. A. (2001). A social identity theory of leadership. Personality and
Social Psychology Review, 5, 184–200.
Hogg, M. A., & Abrams, D. (Eds.). (1988). Social identifications: A
social psychology of intergroup relations and group processes. London:
Routledge.
Hogg, M. A., Hains, S. C., & Mason, I. (1998). Identification and leader-
ship in small groups: Salience, frame of reference, and leader stereo-
typicality effects on leader evaluations. Journal of Personality and
Social Psychology, 75, 1248 –1263.
Hogg, M. A., & Terry, D. J. (2000). Social identity and self-categorization
processes in organizational contexts. Academy of Management Review,
25, 121–140.
Hogg, M. A., & van Knippenberg, D. (2003). Social identity and leadership
processes in groups. In M. P. Zanna (Ed.), Advances in experimental
social psychology (Vol. 35, pp. 1–52). San Diego, CA: Academic Press.
Hollander, E. P. (1958). Conformity, status, and idiosyncracy credit. Psy-
chological Review, 65, 117–127.
Hollander, E. P. (1992). Leadership, followership, self, and others. The
Leadership Quarterly, 3, 43–54.
House, R. J. (1977). A 1976 theory of charismatic leadership. In J. H. Hunt
& L. L. Larson (Eds.), Leadership: The cutting edge (pp. 189 –207).
Carbondale, IL: Southern Illinois University Press.
House, R. J., & Shamir, B. (1993). Toward the integration of transforma-
tional, charismatic and visionary theories of leadership. In M. M. Chem-
ers & R. Ayman (Eds.), Leadership theory and research: Perspectives
and directions (pp. 81–107). New York: Academic Press.
Howell, J. M., & Frost, P. J. (1989). A laboratory study of charismatic
leadership. Organizational Behavior and Human Decision Processes,
43, 243–269.
Hunt, J. H., Boal, K. B., & Dodge, G. E. (1999). The effects of visionary
and crisis-responsive charisma on followers: An experimental examina-
tion of two kinds of charismatic leadership. The Leadership Quarterly,
10, 423– 448.
36
VAN KNIPPENBERG AND VAN KNIPPENBERG
Jacobson, C., & House, R. J. (2001). Dynamics of charismatic leadership:
A process theory, simulation model, and tests. The Leadership Quar-
terly, 12, 75–112.
Kahneman, D. (2003). A psychological perspective on economics. The
American Economic Review, 93, 162–168.
Kark, R., Shamir, B., & Chen, G. (2003). The two faces of transformational
leadership: Empowerment and dependency. Journal of Applied Psychol-
ogy, 88, 246 –255.
Kirkpatrick, S. A., & Locke, E. A. (1996). Direct and indirect effects of
three core charismatic leadership components on performance and atti-
tudes. Journal of Applied Psychology, 81, 36–51.
Lord, R. G., Foti, R. J., & DeVader, C. L. (1984). A test of leadership
categorization theory: Internal structure, information processing, and
leadership perceptions. Organizational Behavior and Human Perfor-
mance, 34, 343–378.
Lord, R. G., & Maher, K. J. (1991). Leadership and information process-
ing: Linking perceptions and performance. Boston, MA: Unwin Hyman.
Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. (1996). Effectiveness
correlations of transformational and transactional leadership: A meta-
analytic review of the MLQ literature. The Leadership Quarterly, 7,
385– 425.
McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detect-
ing interactions and moderator effects. Psychological Bulletin, 114,
376 –390.
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, 1508 –1519.
Platow, M. J., van Knippenberg, D., Haslam, S. A., van Knippenberg, B.,
& Spears, R. (2002, June). A special gift we bestow on you for being
representative of us: Considering leader charisma from a self-
categorization perspective. Paper presented at the 13th General Meeting
of the European Association of Experimental Social Psychology, San
Sebastian, Spain.
Prapavessis, H., & Carron, A. (1992). Sacrifice, cohesion, and conformity
to norms in sport teams. Group Dynamics: Theory, Research, and
Practice, 1, 231–240.
Reicher, S., & Hopkins, N. (2001). Self and nation. London: Sage.
Reicher, S., & Hopkins, N. (2003). On the science of the art of leadership.
In D. van Knippenberg & M. A. Hogg (Eds.), Leadership and power:
Identity processes in groups and organizations (pp. 197–209). London:
Sage.
Rosch, E. (1978). Principles of categorization. In E. Rosch & B. B. Lloyd
(Eds.), Cognition and categorization (pp. 27– 48). Hillsdale, NJ:
Erlbaum.
Rosenthal, R., & Rosnow, R. L. (1985). Contrast analysis: Focused
comparisons in the analysis of variance. New York: Cambridge Uni-
versity Press.
Shamir, B., Arthur, M. B., & House, R. J. (1994). The rhetoric of charis-
matic leadership: A theoretical extension, a case study, and implications
for research. The Leadership Quarterly, 5, 25–42.
Shamir, B., House, R. J., & Arthur, M. B. (1993). The motivation effects
of charismatic leadership: A self-concept based theory. Organization
Science, 4, 577–594.
Shea, C. M., & Howell, J. M. (1999). Charismatic leadership and task
feedback: A laboratory study of their effects on self-efficacy and task
performance. The Leadership Quarterly, 10, 375–396.
Sherif, M., & Sherif, C. W. (1964). Reference groups. New York: Harper
& Row.
Synergistic Learning Incorporated. (1997). Brain works personality soft-
ware. Retrieved July 1997 from http://www.mindmedia.com/interact/
specials/brain.jsp
Tajfel, H. (1982). Social psychology of intergroup relations. Annual Re-
view of Psychology, 33, 1–39.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup
behavior. In S. Worchel & W. G. Austin (Eds.), Psychology of inter-
group relations (pp. 7–24). Chicago: Nelson Hall.
Turner, J. C. (1985). Social categorization and the self-concept: A social
cognitive theory of group behavior. In E. Lawler (Ed.), Advances in
group processes (Vol. 2, pp. 77–122). Greenwich, CT: JAI Press.
Turner, J. C., Hogg, M., Oakes, P. J., Reicher, S., & Wetherell, M. (1987).
Rediscovering the social group: A self-categorization theory. Oxford,
United Kingdom: Blackwell.
van Knippenberg, B., & van Knippenberg, D. (2003). Leadership, identity
and influence: Relational concerns in the use of influence tactics. In D.
van Knippenberg & M. A. Hogg (Eds.), Leadership and power: Identity
processes in groups and organizations (pp. 123–137). London: Sage.
van Knippenberg, D., & Hogg, M. A. (2003). A social identity model of
leadership effectiveness in organizations. In B. M. Staw & R. M. Kramer
(Eds.), Research in Organizational Behavior, 25, 243–295.
van Knippenberg, D., Lossie, N., & Wilke, H. (1994). In-group prototypi-
cality and persuasion: Determinants of heuristic and systematic message
processing. British Journal of Social Psychology, 33, 289–300.
van Knippenberg, D., van Knippenberg, B., & van Dijk, E. (2000). Who
takes the lead in risky decision making? Effects of ingroup members’
individual riskiness, and prototypicality. Organizational Behavior and
Human Decision Processes, 83, 213–234.
van Lange, P. A. M., Rusbult, C. E., Drigotas, S. M., Arriaga, X. B.,
Witcher, B. S., & Cox, C. L. (1997). Willingness to sacrifice in close
relationships. Journal of Personality and Social Psychology, 72, 1373–
1395.
Wofford, J. (1999). Laboratory research on charismatic leadership: Fruitful
or futile. The Leadership Quarterly, 10, 523–529.
Yorges, S. L., Weiss, H. M., & Strickland, O. J. (1999). The effect of leader
outcomes on influence, attributions, and perceptions of charisma. Jour-
nal of Applied Psychology, 84, 428 436.
Yukl, G. (2001). Leadership in organizations (5th ed.). Englewood Cliffs,
NJ: Prentice Hall.
Received April 25, 2003
Revision received January 14, 2004
Accepted January 15, 2004
37
LEADER SELF-SACRIFICE AND PROTOTYPICALITY
... menunjukkan bahwa anggota kelompok yang prototipikal lebih mungkin muncul sebagai pemimpin(Fielding & Hogg, 1997;D. van Knippenberg et al., 2000), dan bahwa pemimpin yang prototipikal lebih berpengaruh dan efektif(Barreto & Hogg, 2017; Barth-Farkas & Vera, 2017;M. J. Platow et al., 2006;M. J. Platow & van Knippenberg, 2001;Steffens et al., 2015;B. van Knippenberg & van Knippenberg, 2005). Prototipikalitas pemimpin dengan demikian merupakan penentu efektivitas kepemimpinan.Prototipikalitas meningkatkan efektivitas pemimpin antara lain karena adanya persepsi bahwa pemimpin yang prototipikal lebih dapat dipercaya dalam mengupayakan kepentingan terbaik kelompok(Barth-Farkas & Vera, 2017;Giessner & van Knippenberg, 2008;Gies ...
... aku pemimpin yang berorientasi kelompok berinteraksi memengaruhi efektivitas kepemimpinan (D.van Knippenberg, 2011).Prediksi bahwa efektivitas pemimpin dipengaruhi perilakunya mendapat dukungan penelitian dalam berbagai contoh perilaku pemimpin yang berorientasi kelompok, seperti pengorbanan diri pemimpin demi kepentingan kelompok(Tee et al., 2014;B. van Knippenberg & van Knippenberg, 2005), keputusan alokasi pelayanan kelompok (M. J.Platow & van Knippenberg, 2001), dan ketertarikan yang lebih tinggi kepada minat kolektif daripada mengikuti kepentingan pribadi (M. J.Platow et al., 2006). M. J. Platow dan vanKnippenberg (2001) mendukung gagasan bahwa efek ini berakar pada identitas sosial dan menegaskan bahwa efek interakti ...
Article
Full-text available
Tujuan studi ini adalah melakukan telaah literatur semi sistematik tentang prototipikalitas pemimpin dalam perspektif identitas sosial. Pencarian menggunakan kata kunci “leader prototypicality” dilakukan dalam berbagai basis data (Emerald, ProQuest, Scopus, Web of Science, dan Portal Garuda) terhadap artikel yang diterbitkan selama dua dekade, dari tahun 2001 hingga 2021. Sebagai hasilnya, diperoleh tujuh puluh satu artikel yang kemudian dianalisis, data diekstraksi dan disentesis sesuai tujuan telaah literatur yang dilakukan, yaitu menjelaskan tentang: (1) pengertian prototipikalitas pemimpin, (2) pengukuran prototipikalitas pemimpin, (3) pengaruh prototipikalitas pemimpin, dan (4) keterbatasan prototipikalitas pemimpin. Pada bagian akhir dilakukan kesimpulan dengan mengajukan saran bagi penelitian kepemimpinan dalam perspektif identitas sosial berikutnya, dengan menggarisbawahi pentingnya kepengikutan.
... For example, perceptions of general leader prototypicality involve subordinates comparing the features of a supervisor (e.g., their appearance, personality, and other attributes) against the set of features that they see as defining leaders in general (Epitropaki & Martin, 2004;Lord et al., 1980). Perceptions of group prototypicality involve similar comparisons against the set of features that they see as defining their shared group (Hogg, 2001;van Knippenberg & van Knippenberg, 2005). Although likely positively correlated with PPPA, the conceptual distinctions between PPPA and these feature-to-prototype comparison processes suggest they will affect endorsement judgments differently, especially in response to the prototype balancing manipulation we test. ...
... Group Prototypicality. We measured participants' perceptions of the degree to which they felt their supervisor was representative of their group prototype using three items adapted from van Knippenberg & van Knippenberg (2005). We asked participants to "please think about the various traits that Captain Jones possesses and how they compare to the traits of other firefighters" and indicate their agreement with the following statements: 1) "Captain Jones is a good example of the kind of people that are firefighters," 2) "Captain Jones represents what is characteristic of firefighters," 3) "Captain Jones has a lot in common with firefighters." ...
Article
Despite decades of efforts, many organizations still have sexist supervisors-those in supervisory positions who define their profession by primarily stereotypically masculine features. As a result of their "masculine" professional prototypes, sexist supervisors see their work as a "man's job" in which women cannot succeed. Research suggests that one problem posed by sexist supervisors is that they may pass their biased views on to subordinates who endorse them as leaders. To make this less likely, we test in two experiments (N = 1,879) a strategy to reduce subordinates' endorsement of sexist supervisors. We do this by encouraging subordinates to see themselves as low in perceived professional prototype alignment (PPPA)-the extent to which a subordinate perceives their supervisor to share their beliefs about what it means to be a member of their profession-with sexist supervisors. Specifically, encouraging subordinates' to hold less masculine, more "balanced" professional prototypes, in which they see stereotypically feminine attributes as equally important to the job as stereotypically masculine ones, reduces PPPA with sexist supervisors. Lowering PPPA, in turn, reduces supervisor endorsement, even after accounting for the effects of other established mechanisms of supervisor endorsement. This research sheds new light on the psychology of followership and offers a new way to curb gender bias from the bottom up. (PsycInfo Database Record (c) 2022 APA, all rights reserved).
... Research shows that potential leaders can become prominent if they can proceed with ingroup goals instead of their own goals (de Cremer, 2002;Haslam et al., 2020;Subašić et al., 2011;van Dick et al., 2018;van Knippenberg & van Knippenberg, 2005;van Vugt & de Cremer, 1999). In the case of populist leadership, the strategy of working against 'the elites', and the anti-political narratives of populist leaders' actions play a key role in effectively embodying this aspect of leadership (Eatwell & Goodwin, 2018). ...
Article
Full-text available
In recent years, the questions of what populism is and how populist leaders mobilize their followers have been the subject of extensive debate. While the social psychology literature holds unique theoretical tools that can be used to explain leader-follower dynamics, these have not yet been applied to understand populism and populist leadership. In this paper, we aim to discuss populism as a social-psychological concept and provide a comprehensive approach to examine the interactions between populist leaders and followers by using the identity leadership model (see New Psychology of Leadership, Haslam et al., 2020). Accordingly , we propose an integrative model in which we suggest that populism should be treated as a social-psychological concept based on (i) strong ingroup identification; (ii) interactive leadership processes that open spaces to followers for enacting their ingroup identity that end up with mobilization against vertical (e.g., elites) and horizontal (e.g., minorities , refugees, opponents) outgroups; (iii) leader's ingroup prototypicality and identity entrepreneurship that is boosted by using shared grievances, narratives of collective victim-hood, and the destabilization of mainstream opponent leaders. Furthermore, by discussing real-world examples and recent studies, we aim to show how the content of what it means to be 'us' and what is seen as moral to 'us' can be shaped by populist leaders for mobilization.
... When the group perceives the leader as representing the group's prototype, they will see him or her as one of them (ingroup), interpret his or her behavior positively and assume he or she is acting in favor of the ingroup (van Knippenberg and Hogg, 2003). Therefore, in organizations, followers (e.g., employees) tend to like and trust a prototypical leader more than non-prototypical leaders, they are more likely to tolerate the shortcomings of a prototypical leader and ascribe them a higher degree of leadership effectiveness (van Knippenberg and Hogg, 2003;van Knippenberg and van Knippenberg, 2005;Giessner and van Knippenberg, 2008;van Dijke and De Cremer, 2008;Giessner et al., 2009;Ullrich et al., 2009;Steffens et al., 2021). Two recent meta-analyses have confirmed the positive effects of leader group prototypicality for a range of individual and organizational constructs (Barreto and Hogg, 2017;Steffens et al., 2021). ...
Article
Full-text available
In the present study, we complement role congruity theory with insights from the Social Identity Model of Leadership. We propose that especially female leaders benefit from team prototypicality, i.e., being representative of the group they are leading. We assume that team prototypicality shifts the comparative frame away from higher-order categories like gender and leader roles to more concrete team-related properties and thereby reduces disadvantages for female leader that stem from the incongruity between the leader role and the female gender role stereotypes. Further, this effect should affect both (female) leaders themselves and their perception by their followers. Building on previous research, we predict, first, lower authentic leadership behavior for female than male leaders. Second, that team prototypicality positively relates to authentic leadership and trust in leader. Third, that team prototypicality has stronger relations to authentic leadership and trust in leader for female compared to male leaders. We tested assumptions in a randomized online experiment (Study 1, N = 315) and a cross-sectional survey study (Study 2, N = 300). We did not find consistent support for the assumed gender differences in authentic leadership. But our results (both in manifest and in latent analyses) show that team prototypicality-both self-perceived (Study 1) and as perceived by employees (Study 2)-is related to more authentic leadership and more trust in leader (Study 2) and that these relations are stronger for female than for male leaders. Furthermore, we tested in Study 2 an extended model including follower's job satisfaction as the final follower outcome affected via team prototypicality, leader gender, authentic leadership, and trust in leader. Thereby, we found that team prototypicality has direct and indirect effects on job satisfaction as carried through authentic leadership and trust in leader, respectively. Together, the results of both studies support our assumptions and show that female leaders can reduce role incongruity barriers through high team prototypicality. Implications for future research and practical implications of these results for gender equality are discussed.
... can in turn lead to improved work relationships and likely, a more positive self-image and reputation for the one enacting CS. Moreover, showing willingness and even more the actual CS behavior in the context of one's team, such as leadership or team member role related sacrifice, will provide a powerful signal/role modeling for the followers and colleagues in general (Van Knippenberg and Van Knippenberg, 2005). CS behavior in the context of the leadership role may also set a norm for others and show dedication to a bigger cause or to a particular organization. ...
Article
Full-text available
In the ever more uncertain career context, many individuals engage in a form of career sacrifice (CS) at some point in their career journey; that is, giving up of certain career goals/actions or reshaping career decisions to accommodate specific work or life demands. This conceptual paper unpacks CS as an important yet little explored dimension of career decision making. Specifically, the paper examines possible triggers of CS as well as the diverse nature of CS, ranging from short-term (usually minor) type of sacrifice to more significant and long-term sacrifice. We explore the context of this type of career decision making, specifically the intersection of work and non-work-related triggers and conclude by discussing possible work and non-work outcomes both at the individual as well as organizational level. CS outcomes range from enhanced career self-management and relational benefits to positive organizational contributions, but at times can also lead to regret. Areas for future research are identified, especially exploration of demographic and more macro level variables as possible moderators in CS decisions. Future theoretical development of CS is discussed too.
Article
Love (in the agape form) forms the foundation of most leadership concepts and has been ignored in research. We respond to the debate on universal applicability of leadership forms by bringing followers into the spotlight through our examination of the interactive influence of loving (agape-based) and non-loving (non-agape-based) leadership styles and followers’ attachment dimensions (self-model and other-model) on follower outcomes. Two hundred and eighty-two business management students worked in teams on a task under the direction of leaders who demonstrated agape-based behaviours and leaders who demonstrated non-agape-based behaviours in a laboratory experiment. Agape-based leadership was positively related with follower satisfaction with the leader, team commitment and perception of leaders’ effectiveness. Further, followers’ attachment dimensions (self- and other-model) moderated the relationship between agape-based leadership and follower work attitudes, such that the relationship was positive for followers with a negative self-model and for followers with a positive other-model, and the relationship was negative for followers with a negative other-model. We provide a practical set of tools for demonstrating agape leadership behaviours which are useful for educators and organizations. We suggest that leaders must alter their leadership style depending on their followers’ attachment dimensions.
Article
I contend that in leadership research, research method preferences guide the development of research questions rather than the other way around as would prescriptively be expected. This is a problem because leadership research is heavily dominated by survey research. This low diversity in research method preference constrains the consideration set of research questions: leadership research tends to focus on what can be studied in surveys, not because these are necessarily the most interesting issues to study but because of a preference for survey research. This dominance of survey research is self-sustaining both because people are more likely to use the methods they are more familiar with and because it creates an implicit norm as to what are appropriate research methods and appropriate research questions. I discuss how stimulating diversity in research methods is needed to address this problem and to let diversity in research questions follow.
Article
As organizational research continues to globalize, scholars increasingly must translate established scales into languages other than those in which the scales were originally developed. In organizational psychology research, back-translation is the dominant procedure for translating scales. Back-translation has notable strengths in maintaining the psychometric properties of an established scale in a translated version. However, cross-cultural methodologists have argued that in its most basic form, back-translation often does not result in translations with acceptable levels of equivalence between original and translated research materials. Fortunately, there are complementary procedures to back-translation that can evaluate and strengthen the extent to which scale translations have achieved equivalence between original and translated versions of scales. But how often organizational researchers use and report these procedures in tandem with back-translation is unclear. This article aims to address this lack of clarity by evaluating the state of the use of back-translation in organizational psychology research by reviewing every study in Journal of Applied Psychology that has employed translation over the past nearly 25 years (k = 333). Our findings suggest that the majority of the time that researchers engage in translation procedures, they report having done so. At the same time, the details of these procedures are commonly underreported, making it unclear whether additional techniques beyond back-translation have been used to examine and demonstrate equivalence between original and translated versions of scales. Based on the results of our review, we develop a set of recommendations for conducting and reporting scale translations in organizational research. (PsycInfo Database Record (c) 2022 APA, all rights reserved).
Article
Full-text available
Previous research has predominantly regarded self-sacrificing leaders as role models that can drive desirable outcomes. We challenge this notion and demonstrate the dual effects of self-sacrificial leadership on team prosocial and change-oriented behaviors. Drawing upon social learning theory, we develop a nuanced model that simultaneously examines self-sacrificial leadership's beneficial effects on team helping and its detrimental effects on team change-oriented behavior via unique mechanisms. Our analysis of two time-lagged, multisource data revealed that self-sacrificial leadership has an indirect positive effect on team helping via team other-focus and an indirect negative effect on team change-oriented behaviors (i.e., team creativity and team proactivity) via team dependence on the leader. Furthermore, team leader competence amplifies these indirect effects, such that the effects are stronger for more competent leaders. Taken together, our research challenges the prevailing assumption that self-sacrificial leadership is always beneficial and calls for further attention to be paid to the unintended impacts of role model leaders.
Article
Purpose The purpose of this paper was to lay the necessary conceptual and empirical groundwork of agape in organizations. Specifically, the authors reviewed literature on agape; advanced formal definition of agape; explained the relationship of agape with related variables; developed a scale to measure agape and provided evidence of its reliability and construct validity; showed how agape uniquely predicted employee outcomes beyond transformational leadership; and showed how agape compensated for the lack of transformational leadership. Design/methodology/approach The authors conducted a survey with 214 working executives who rated their manager on transformational leadership and agape behaviours, and later indicated their own work attitudes. Next, the authors conducted a 20-min between-subjects vignette experiment with 147 business management students who were provided with a description of a supervisor and asked to indicate their work attitudes under the supervisor. Findings The authors advanced an operational definition and a scale to measure agape. The findings of this study indicated that agape was a unidimensional construct with high reliability. It had significant positive relationships with followers’ job satisfaction, faith and loyalty, team commitment, satisfaction and risk-taking; explained incremental variance in employee outcomes beyond transformational leadership; and compensated for the lack of transformational leadership. Research limitations/implications The present research has the potential to inform recruitment, selection, training, promotion and performance evaluation decisions in organizations. Originality/value The authors responded to calls for developing a clear and consistent conceptualization and operationalization of agape for improving scholarly research and leadership training and development.
Article
offers a comprehensive description of the development and validation of transformational leadership theory / in response to criticisms of the conceptualization, measurement, and evidential bases of the theory, the authors bring together the results of an impressively extensive program of research / they identify what they consider to be both the strengths of their approach as well as the areas needing further development / offer a future agenda for research and training (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Chapter
Leadership, identity and influence: Relational concerns in the use of influence tactics The study of leadership and the study of influence are closely intertwined: the two topics are often discussed concurrently in handbooks and review articles and they are frequently treated as more or less similar or related concepts in empirical studies. Some researchers make an even stronger link between leadership and influence. They claim that the use of influence is the essence of leadership (e.g., Yukl, 1998). After all, leaders need to guide, structure, mobilize, facilitate, envision, and define identity, foster harmonious relations, and enhance performance (in organizations or other groups), and they can only do so by exercising influence. How leaders exercise influence is therefore a core question for students of leadership. In this chapter, we use theoretical insights in identity processes to present a framework to understand the choice and effectiveness of leaders’ influence attempts. Building on ...
Chapter
On the science of the art of leadership No area of modern social thought has escaped the shadow of the holocaust. The issues that we prioritize, the questions that we ask and the perspectives that we employ all changed irrevocably as a result of the slaughter. Leadership research is a case in point. Prior to the Second World War, many thinkers were fascinated and attracted by forceful charismatic leaders who were seen as saving society from a dull mechanical future. Such figures stamped some agency, artistry and imagination on what Max Weber described as ‘the routinized economic cosmos’ (quoted in Lindholm, 1990, p. 27). However, after Hitler, ‘the triumph of the will’ acquired different connotations. The focus shifted from what was added by the leader to what was taken away from everyone else and hence the savior became Satan. Thus, in the post-war period, there were a plethora of clinical ...
Chapter
The vast body of theory and research on justice in groups and organizations has focused primarily on issues of distributive justice (Homans, 1961), that is, the manner in which resources are distributed, and on responses to these distributions (for recent reviews, see Freedman & Montanari, 1980; Greenberg, 1982). Another fundamental type of justice manifest in groups and organizations, but one that has received considerably less attention, concerns the rules and processes through which resources are allocated, that is, procedural justice (Leventhal, 1976; Thibaut & Walker, 1975; Tyler & Caine, 1981). For example, in considering the fairness of pay raises, workers may not only take into account how much pay they receive relative to others, but also such procedural factors as who made the decision, and what criteria were taken into account (see Lawler, 1971). Concerns of this type, focusing on the process of allocation, rather than on the outcome of allocation per se, fall into the domain of procedural justice.