Professional Documents
Culture Documents
net/publication/228338351
CITATIONS READS
149 6,239
4 authors, including:
Michael Grojean
University of Tennessee
9 PUBLICATIONS 1,388 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Rolf van Dick on 14 May 2019.
The
British
Psychological
Journal of Occupational and Organizational Psychology (2007), 80, 133–150
q 2007 The British Psychological Society
Society
www.bpsjournals.co.uk
We present a multi-sample multi-level approach that examines the link between leader
and follower organizational identification, and follower attitudes. Study 1 comprises
367 school teachers and 60 head teachers in Germany. The results illustrate a significant
relationship between head teacher and school teacher school identification. Moreover,
indirect relations between head teacher school identification and school teacher job
satisfaction and self-reported citizenship behaviours, mediated by school teacher
school identification, are predicted and supported by the data. The findings are
replicated within Study 2, comprising 233 school teachers and 22 head teachers. Finally,
a third study replicates the findings in a different sector using a sample of 314 travel
agents in 127 travel agencies and their leaders. Taken together, leader’s self-construal in
terms of the organization is related to follower organizational identification, and
therefore leads to greater follower satisfaction and to a greater willingness to exert
extra effort on behalf of the organization.
As organizations move forward into the twenty-first century, the nature of business is
changing. Companies are becoming smaller, hierarchical structures are giving way to a
wider variety of forms, many businesses are shifting emphasis from production to
service and work itself is being redefined along lines of improved effectiveness (Hammer
& Champy, 1993; Kantner, Stein, & Jick, 1992). In conjunction with these changes, a
growing sense of the importance of people in organizations is developing (Doyle, 1990).
The development and expansion of such ‘people power’ has become the focus of
organizational efforts, such that human resource programmes are increasingly
called upon to enhance organizational performance and employee commitment
(Conway, 2004). Grojean and Thomas (2005) extend this argument to suggest that the
* Correspondence should be addressed to Rolf Van Dick, Goethe University Frankfurt, Institute of Psychology, Kettenhofweg
128, 60054 Frankfurt, Germany (e-mail: van.dick@psych.uni-frankfurt.de).
DOI:10.1348/096317905X71831
134 Rolf Van Dick et al.
for the nested structure of the data. In the following section, we review the literature
and the hypotheses comprising this framework.
Hypothesis 2. Leader organizational identification will relate to follower job satisfaction and
OCB; the more leaders identify with their organization, the more follower satisfaction and OCB
is expected.
(Mael & Ashforth, 1992; Van Knippenberg, 2000). This in turn strengthens work
motivation and ultimately performance. In-role performance is determined by several
factors, of which only some are under the volitional control of the individual. Thus, the
expected positive effect of identification on performance should be most prevalent for
forms of extra-role or citizenship behaviours. Accordingly, research has consistently
found that individuals who are strongly identified with their teams and/or organizations
are more satisfied and more likely to show extra-role behaviour (Riketta & Van Dick,
2005). A recent meta-analysis, for instance, has revealed average correlations between
identification and satisfaction of r ¼ :54, and between identification and extra-role
behaviour of r ¼ :35 (Riketta, 2005). Consistent with these findings, we put forward the
following hypothesis:
Hypothesis 4. The relationship between leader organizational identification and follower job
satisfaction and OCB will be mediated by follower organizational identification.
138 Rolf Van Dick et al.
STUDY 1
Method
Sample and procedure
The original sample consisted of 515 female (N ¼ 316) and male (N ¼ 199) teachers,
stemming from all German school types and from eight different federal states. Mean age
was 46 years (SD ¼ 9 years), mean tenure was 19.4 years (SD ¼ 10:6 years).
Questionnaires were distributed by student research assistants in 86 different schools.
Questionnaires were provided to all teachers via their pigeon-holes and, upon
completion, put in sealed envelopes and then collected by the schools’ administrative
staff. Overall response rate was 63%. Here, a subsample of N ¼ 367 teachers from
60 schools will be used for the analyses where information was also collected among the
schools’ head teachers. Twenty-one of the head teachers were female (35%), and the
mean age was 53 years (SD ¼ 6:4 years). Head teachers were in their current position for
an average of 11 years (SD ¼ 7 years).
Questionnaire
The head teachers’ questionnaire comprised a single statement asking the head teachers
to rate their identification with their school (‘I am personally identified with my
school’). The answering format for this item and all items used in the school teachers’
questionnaires was a six-point scale with end-points not at all true and perfectly true,
respectively. We used this measure because time limitations in the head teachers’
questionnaire permitted the use of this item only.
Teachers rated identification with their school using a 6-item measure developed by
Van Dick, Wagner, Stellmacher, and Christ (2004; e.g. ‘being a member of this school is
an important reflection of who I am’). Job satisfaction was measured with 3 items from
the job diagnostic survey (Hackman & Oldham, 1980). OCB was measured using a
6-item scale collating two dimensions of OCB towards individuals (altruism) and
towards the organization (conscientiousness) as described by Christ, Van Dick, Wagner,
and Stellmacher (2003, e.g. ‘I participate actively in meetings and conferences’, ‘to
questions or uncertainties of colleagues, I help always readily’). Cronbach alphas for the
three scales, respectively, were a ¼ :80, .70 and .81. Table 1 provides intercorrelations
and descriptive statistics for each scale.
Teachers
Identification 4.33 1.04 .80 .40** .45**
Job satisfaction 5.11 0.85 .70 .31**
OCB 4.32 0.82 .81
Head teachers’
Identification 5.59 0.95a .14b .14b 2.01b
Results
The data gathered for this study are of a nested nature. There are potentially two ways of
addressing the fact that we have a number of followers associated with each leader. One
technique would be to aggregate follower information (i.e. identification, satisfaction,
OCB) and to correlate this aggregated information with the data gathered from head
teachers. There are, however, two problems. One problem is that by aggregating
information of 360 individuals to 60 group-level scores, one loses statistical power. The
more serious problem, however, lies in the fact that aggregation is only justified if the
concepts under study are conceptualized on the group level. Clearly, teacher
identification, satisfaction and self-reported OCBs are individual-level constructs and
should thus be treated as such. We thus conducted our analyses using hierarchical linear
modelling (HLM) employing the MLwiN (Version 2.0) software. The strategy involved
expressing the individual-level outcome satisfactionij (i.e. the satisfaction of the ith
teacher in the jth school) using a pair of linked models – one at the level of the individual
school teachers and one of the level of the head teachers. In the individual-level model,
satisfaction of the ith teacher in the jth school was expressed as the sum of an intercept
for the teacher’s school (b0j) and random error (rij) associated with the ith teacher in the
jth school. In the group-level model, individual-level intercepts were expressed as the
sum of an overall mean (g00), a fixed effect for head teacher identification (g01), and a
series of random deviations (m0j).Substituting the group-level model yields the multi-
level model (e.g. satisfactionij ¼ g00 þ g01 head teacher identificationj þ m0j þ rij).
This multi-level model tests whether the satisfaction of the ith teacher in the jth school
can be predicted from the identification of that teacher’s school head teacher
(and accordingly in the analyses for school teacher OCB and own identification).
All analyses were conducted using type of school as a control variable. Teachers in
this study were employed in seven different types of school (primary school, grammar
school, three German versions of secondary comprehensive school, sixth form schools
and special needs school), which we coded using six dummy variables in the analyses.
We included this control variable to account for school type differences associated with
school size and status (i.e. grammar schools are usually larger and more prestigious than,
for example, primary schools or special needs schools). Thus, controlling for school
type takes into account the possibility that head teacher and school teacher
identification reflect their commitment to working in a school that is, for instance,
larger, more prestigious or better resourced. First, we tested whether schools differed
with respect to the two dependent variables. For this purpose, we tested between-
group variation of the random effect components. For both job satisfaction (m0j ¼ :063,
SE ¼ :032; x2 ¼ 3:84, p , :05) and OCB (m0j ¼ :064, SE ¼ :031; x2 ¼ 4:23, p , :05),
the respective terms were substantial and significant. For the following mediation
analyses, we will basically follow the procedure suggested by Baron and Kenny (1986).
That is, we will first test for a relationship between predictor (leader identification) and
the outcomes (follower satisfaction and OCB, respectively). In the second step, we will
test the relationship between predictor and mediator (follower identification), and in
the third step, we will test the relationship between mediator and outcomes. In the
fourth step, following Baron and Kenny, we will test whether the relationship between
predictor and outcome shrinks upon the addition of the mediator to the model. These
criteria can be used to estimate whether or not mediation is occurring, but MacKinnon
and colleagues (MacKinnon & Dwyer 1993; MacKinnon, Warsi, & Dwyer, 1995) have
developed methods by which mediation can be formally assessed. Thus, in an additional
140 Rolf Van Dick et al.
and final step, we will apply the Sobel procedure (Sobel, 1982) for testing significance of
the indirect effect (see Preacher & Leonardelli, 2003).
There is an ongoing debate, however, whether all steps of Baron and Kenny’s model
are necessary for tests of mediation. More specifically, James, Mulaik, and Brett (2005)
recently argued that the correlation tested in the first step of Baron and Kenny’s model
(i.e. the direct relationships between independent and dependent variable) does not
always need to be significant to meet the criteria of mediation (i.e. an indirect effect of
the independent on the dependent variable through the mediator). In fact, Kenny,
Kashy, and Bolger (1998) have proposed that situations exist in which one might look
for evidence of mediation in the absence of a relation between predictor and outcome.
Shrout and Bolger (2002; see Frazier, Tix, & Barron, 2004), for instance, suggested
skipping the first step if the predictor is distal to the outcome (e.g. in longitudinal
studies) because of low power to detect direct predictor outcome. Following this
reasoning, we will test for effects of the mediator (i.e. follower identification) in all
following analyses, whether or not a direct relationship between the predictor (leader
identification) and outcomes (follower satisfaction and OCB) can be established.
Consistent with Hypothesis 1, head teacher school identification was significantly
related to followers’ levels of school identification (b ¼ 0:16, p , :05).
Partial support was provided for Hypothesis 2. Head teacher identification was
unrelated to teacher OCB (b ¼ 0:03, ns) but related to teacher satisfaction (b ¼ 0:10,
p , :05, one-tailed). Hypothesis 3 was confirmed as teacher identification was
significantly related to their satisfaction (b ¼ 0:44, p , :01) and self-reported OCB
(b ¼ 0:46, p , :01).
With support for Hypotheses 1 and 3, two of the three conditions to test mediation
were met. Head teacher identification was not related to OCB but was related to job
satisfaction. The relationship between head teacher identification and teacher
satisfaction decreased to an insignificant b ¼ 0:03 after including the mediator. For
OCB, we followed the suggestions of Shrout and Bolger (2002; see discussion above)
and applied the Sobel procedure (Sobel, 1982) to test whether head teacher
identification was indirectly related to teacher satisfaction and OCB. The Sobel
procedure indeed revealed such indirect effects via school teacher identification for
teacher satisfaction, tð1Þ ¼ 2:45, p ¼ :014, and OCB, tð1Þ ¼ 2:49, p ¼ :013. Thus,
although we could not establish direct relationships between head teacher
identification and school teacher OCB, there was evidence for a mediation effect of
follower self-identity as proposed by Lord and Brown (2004) and Van Knippenberg et al.
(2004).
Discussion
The main hypotheses of Study 1 have been confirmed. Leader identification seems to
have a contagious effect on follower identification, which, in turn, is related to the
follower attitudes and self-reported behaviours. The present sample was reasonably
large and scales for assessing follower attitudes and OCBs revealed sound statistical
properties. We will discuss issues of the cross-sectional nature of our data in the general
discussion of this paper as this problem affects all studies in a similar way. One specific
problem of the present study, however, needs to be mentioned here. Of course,
assessing head teacher identification with a single item is difficult. The study was not
designed to test cross-level interactions and thus no attention has been paid to using
identical items and scales for teachers and head teachers. Thus, the study suffers from
Social identification from a leader-follower perspective 141
the fact that for administrative reasons the head teacher questionnaire had to be as short
as possible. The single item used for assessing head teacher identification (‘I am
personally identified with my school’) is the most valid item we can imagine but, of
course, no estimates of reliability and consistency can be provided. Furthermore, the
high mean for this item and its skewness (in fact, 76% of all head teachers fully agreed to
the statement) point to some problems of the item difficulty. To counter this limitation,
we developed and administered the second study, the aim of which was to replicate the
findings of Study 1 while using a more sophisticated measure with respect to head
teacher identification.
Despite the limitations discussed, having shown for the first time that leader
identification is related to follower identification and in turn follower attitudes is a
critical first step demonstrating the usefulness of our theoretical reasoning. The
question remains why head teacher and school teacher identification is not related more
strongly as both share the same environment and may have similar perceptions of what
their school’s distinctive and enduring features were. The relatively small relationship
obtained in the cross-level hierarchical model shows that only 5% of variation between
head teacher and school teacher identification is shared. This again may be due to the
use of a single item for head teacher identification and shall be fully discussed in relation
to Studies 2 and 3 at the end of this paper.
STUDY 2
Method
Sample and procedure
The original sample consisted of 464 teachers from two different German school types
(primary school, N ¼ 195, and secondary high school, N ¼ 257) and four different
federal states. The aim of this study was to test effects of manipulating the salience of
different focuses of identification (career, school, occupation; cf. Van Dick, Wagner,
Stellmacher, & Christ, 2005). However, a control group of 233 teachers within the total
sample received questionnaires with no manipulation at all, and this control group is the
basis for the following analyses. Fifty-eight percent of the sample was female, with a
mean age of 46 years (SD ¼ 9:8 years), and a mean tenure of 18.8 years (SD ¼ 10:4
years). Questionnaires were distributed by student research assistants in 22 different
schools. Questionnaires were provided to all teachers via their pigeon-holes and, upon
completion, put in sealed envelopes and then collected by the schools’ administrative
staff. The overall response rate was 45%. Half of these 22 school’s head teachers were
female (50%), and mean age was 53 years (SD ¼ 6:6 years). Head teachers were in their
current position for an average of 11 years (SD ¼ 7:8 years).
Questionnaire
School Identification was obtained using the same instrument (Van Dick et al., 2004) as
in the teachers’ questionnaire of Study 1. This time, six identical items were used for
assessing identification of both teachers and head teachers. Reliability coefficients were
good (a ¼ :81 for teachers, and a ¼ :72 for head teachers, respectively). Satisfaction
was assessed with a single item (‘I am glad that I work at this school’). This item stems
from a survey tool that has been developed for organizational diagnosis in schools
(Ulber, 2001).
142 Rolf Van Dick et al.
Citizenship behaviours were measured with 5 items specifically formulated for the
teaching profession (cf. Christ et al., 2003; e.g. ‘I frequently make innovative
suggestions in meetings or towards the principal’), and the scale had a reliability of
a ¼ :77. Table 2 provides intercorrelations and descriptive statistics for each scale. The
answering format for all items was a 6-point scale with end-points not at all true, and
perfectly true, respectively.
Teachers
Identification 4.40 0.97 .81 .53** .46**
Job satisfaction 5.10 1.17a .33**
OCB 4.26 1.03 .77
Head teachers’
Identification 5.23 0.61 .72 .16b .23b .03b
Results
All analyses were again conducted using hierarchical linear modelling, and again we
controlled for school type (in this case, primary versus secondary schools as a
dichotomous variable) as in Study 1. School type was related to school size (number of
students) with r ¼ :87 (p , :001), so controlling for type also controls for size effects
but incorporates other factors like differences in resources etc. that may have effects on
teacher and head teacher identification. First, we tested whether schools differed with
respect to the dependent variables of school satisfaction and OCB. For this purpose, we
tested between-group variation of the random effect components. The variation was
significant for school satisfaction (m0j ¼ :32, SE ¼ :103; x2 ¼ 4:64, p , :01) but not for
OCB (m0j ¼ :08, SE ¼ :054; x2 ¼ 2:19, ns). For the following mediation analyses, we
used the procedures described in Study 1.
Consistent with Hypothesis 1, head teacher school identification was significantly
related to follower school identification (b ¼ 0:13, p , :05).
As predicted in Hypothesis 2, head teacher school identification was significantly
related to teacher satisfaction (b ¼ 0:15, p , :05). There was, however, no significant
relationship between head teacher school identification and teacher OCB (b ¼ 0:04,
ns). Hypothesis 3 was confirmed as teacher identification was significantly related to
their satisfaction (b ¼ 0:51, p , :01) and OCB (b ¼ 0:37, p , :01). The effect of head
teacher identification on school teacher school satisfaction was mediated as predicted in
Hypothesis 4 and did not remain significant after including follower identification
(b ¼ 0:08, ns).We applied the Sobel procedure again and found significant indirect
effects of head teacher identification on teacher satisfaction, tð1Þ ¼ 2:00, p ¼ :04, via
school teacher identification.
Again, as in Study 1, head teacher identification was not directly related to teacher
OCB. Thus, we followed the suggestions of Shrout and Bolger (2002) and applied the
Sobel procedure (Sobel, 1982) to test whether head teacher identification was indirectly
Social identification from a leader-follower perspective 143
related to teacher OCB. The Sobel procedure indeed revealed such indirect effects via
school teacher identification for teacher satisfaction and OCB, tð1Þ ¼ 2:01, p ¼ :044.
Thus, although we could not establish direct relationships between head teacher
identification and school teacher OCB, there was evidence for a mediation effect of
follower self-identity.
Discussion
The general aim of Study 2 was to replicate the findings of our first study in a
different sample but within the same occupational group. Largely, our results have
been confirmed with significant relationships between leader and follower
identification and a mediation effect of follower identification between leader
identification and follower attitudes and behaviours. This has been achieved using
identical scales for both school teacher and head teacher identification, thus
rectifying the main weakness of Study 1. This study, however, suffers from the fact
that we only used a single item assessing job satisfaction and from a relatively low
sample size. The aim of the third study thus is to replicate the findings in a bigger
sample. A secondary aim of this study is to replicate the findings in an organizational
setting that is different from public schools and to test our hypotheses in the private
sector service industry.
STUDY 3
Method
Sample and procedure
The original sample consisted of 388 travel agents from travel agencies in 15 different
German federal states. Of the sample, 90% was female, and mean age was 31 years
(SD ¼ 8:6 years), with mean tenure 6.3 years (SD ¼ 5:6 years). Questionnaires were
distributed by student research assistants in 152 different travel agencies. Separate
return envelopes were provided for each respondent and questionnaires were
returned to the researchers via postal mail. The overall response rate was 48%. Here,
a subsample of 314 travel agents from those 127 travel agencies will be used for the
analyses where information was also collected among the travel agency managers. Of
the leaders, 50% were female, and mean age was 40 years (SD ¼ 9:9 years).
Managers were working at their travel agency for an average of 14 years (SD ¼ 9:5
years).
Questionnaire
Organizational Identification was measured using Mael and Ashforth’s (1992) 6-item
scale (e.g. ‘When someone praises this travel agency, it feels like a personal
compliment’). Reliability coefficients were good (a ¼ :89 for travel agents, and a ¼ :86
for managers, respectively). Organizational citizenship was assessed with three items
(e.g. ‘I inform others about current developments in the organization’). Reliability was
good with a ¼ :79. Table 3 provides intercorrelations and descriptive statistics for each
scale. The answering format for all items was a 7-point scale with end-points not at all
true and perfectly true, respectively.
144 Rolf Van Dick et al.
Travel agents
Identification 3.54 1.29 .89 .34**
OCB 3.51 1.31 .79
Managers
Identification 3.04 1.20 .86 .15a .17a
Notes.**p , :01.
a
Cross-level correlations calculated on individual level, significance testing not applicable.
Results
All analyses were conducted controlling for number of employees and travel agency’s
financial performance (profit before tax as a percentage of sales before tax). Neither of
the controls had effects on follower identification (for number of employees:
B ¼ 20:062, SE ¼ :089; for financial performance: B ¼ 20:036, SE ¼ :086). Again, we
used hierarchical linear modelling to test our hypotheses. First, we tested whether travel
agencies differed with respect to the dependent variable of OCB. For this purpose, we
tested between-group variation of the random effect components. The variation was
significant (m0j ¼ :259, SE ¼ :117; x2 ¼ 4:92, p , :05). For the following mediation
analyses, we followed the procedures described in Study 1.
Consistent with Hypothesis 1, manager organizational identification was significantly
related to follower organizational identification (b ¼ 0:19, p , :05).
As predicted in Hypothesis 2, manager identification was significantly related to
travel agents’ OCB (b ¼ 0:18, p , :05). Hypothesis 3 was confirmed as travel agent
identification was significantly related to their OCB (b ¼ 0:36, p , :01). The effect of
manager identification on travel agents’ OCB was mediated as predicted in Hypothesis 4
and did not remain significant after including follower identification (b ¼ 0:13, ns).
We applied the Sobel procedure again and found significant indirect effects of
manager identification on travel agents’ OCB, tð1Þ ¼ 2:09, p ¼ :04, via travel agent
identification.
Discussion
Study 3 replicates the findings from the first two studies with respect to one dependent
variable, OCB. Unfortunately, Study 3 did not include a measure of job satisfaction, as the
study was not originally designed as a replication and was serving mainly other
purposes. However, the replication regarding OCB advances beyond the first two
studies by focusing on a different sector and thus lends support to the assumption that
our theoretical propositions and findings can be generalized.
GENERAL DISCUSSION
This research set out to test whether leader organizational identification was related to
follower organizational identification and in turn to follower satisfaction and OCB.
Consistent with Hypothesis 1, leader organizational identification was significantly
related to follower organizational identification in all three samples. Partial support was
provided for Hypothesis 2. Leader organizational identification was directly related to
Social identification from a leader-follower perspective 145
follower satisfaction in Studies 1 and 2, and to OCB in Study 3. For OCB in Studies 1 and
2, a direct relationship could not be established. Hypothesis 3 was supported for job
satisfaction in Studies 1 and 2 and for OCB in all three samples. Follower organizational
identification was significantly related to satisfaction and self-reported OCB. Support
was also provided for Hypothesis 4; follower organizational identification mediated the
relationship between leader organizational identification and follower satisfaction and
OCB.
Consistent with Hypothesis 1, leader organizational identification was related to
follower organizational identification. We ascribe this effect to contagion of the leader
identification to the follower identities because of the leaders’ behaviours depending on
his or her own identity. There are, of course, other possibilities that may be responsible
for the leaders and followers sharing identities such as the organization’s reputation or
economic condition. Therefore, it is very important to note that we found the
hypothesized effects, even when controlling for external variables such as school type in
Studies 1 and 2 and travel agencies’ size and financial performance in Study 3.
Combining the present findings and experimental evidence showing that leader group-
oriented behaviour increases follower identification (De Cremer & Van Knippenberg,
2002), we conclude that leaders contribute to follower collective identity the more they
are identified with their organization themselves.
It is interesting, however, that leader and follower identification are only weakly
correlated across all three studies and share a maximum of 5% of variation. This finding
is a bit puzzling as head teachers and school teachers, and travel agencies’ leaders and
agents, respectively, share the same environment. The effect, however, makes sense if
one takes into account that leaders and followers can take very different perspectives on
the organization. Leaders act as boundary spanners and communicate with different
entities (other schools, other agencies, corporate headquarters or departments of
education etc.) more than followers, who have a more inward perspective and
communicate, for instance, with other teams within the organization. This may render
identifications within the organization more salient for followers, such as identification
with their teams or work-groups, whereas leaders are more inclined to identify with the
organization as a whole as this is the more salient frame of reference on their level (see
Riketta & Van Dick, 2005).
Partial support was provided for Hypothesis 2. Leader identification was related to
follower satisfaction in Studies 1 and 2 and to OCB in Study 3 but only indirectly to OCB
in Studies 1 and 2. We have argued that, as the concepts are relatively distal in nature,
proof of the indirect relationships sufficiently supports our basic hypothesis that leader
identification can translate into follower attitudes and behaviours. Thus, our basic
rationale of follower identification as the main mechanism underlying leaders’ impact
on follower attitudes and behaviours is largely confirmed.
Consistent with Hypothesis 3, follower identification was significantly related to
OCB across the three studies and to satisfaction in Studies 1 and 2. As the data are cross-
sectional and based on self-reports, this may not be surprising and should not be
overestimated.
Practical implications
This research extends our knowledge of how a leader’s identification with a collective
can have a contagious effect on subordinates. These results also suggest that trends
which act to reduce individuals’ identification with their organization (e.g. reduced job
146 Rolf Van Dick et al.
Acknowledgements
Empirical studies were supported by the German Science Foundation (DFG; WA 1150/5-1; WI
3146/1) and by a travel grant to the first author from the British Academy (OCG-40013).
The sample of Study 1 was used by Van Dick et al. (2004) and the sample of Study 2 was used by
Van Dick et al. (2005) but leadership issues and head-teachers’ information have not been subject
to the previous analyses
Parts of this paper were presented at the annual meeting of the Society for Industrial and
Organizational Psychology (SIOP), Los Angeles, April, 14-17, 2005, and at the European
Association of Experimental Social Psychology (EAESP) meeting on Progress in Social Identity
Theory, Exeter, July, 13-15, 2005.
We are grateful to John Arnold, Jeremy Dawson, Michael Hogg, Andreas Richter, Daan Van
Knippenberg, and two anonymous reviewers for their helpful comments on earlier versions of this
paper.
References
Baron, R. M., & Kenny, D. A. (1986). The moderator mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173–1182.
Baron, R. S., Crawley, K., & Paulina, D. (2003). Aberrations of power: Leadership in totalist groups.
In D. Van Knippenberg & M. A. Hogg (Eds.), Leadership and power (pp. 169–183). London:
Sage.
Christ, O., Van Dick, R., Wagner, U., & Stellmacher, J. (2003). When teachers go the extra-mile:
Foci of organisational identification as determinants of different forms of organisational
148 Rolf Van Dick et al.
Lord, R. G., & Brown, D. J. (2004). Leadership processes and follower self-identity. Mahwah, NJ:
Erlbaum.
MacKinnon, D. P., & Dwyer, J. H. (1993). Estimating mediated effects in prevention studies.
Evaluation Review, 17, 144–158.
MacKinnon, D. P., Warsi, G., & Dwyer, J. H. (1995). A simulation study of mediated effect measures.
Multivariate Behavioral Research, 30, 41–62.
Mael, F., & Ashforth, B. E. (1992). Alumni and their alma mater: A partial test of the reformulated
model of organizational identification. Journal of Organizational Behavior, 13, 103–123.
Moreland, R. L., & Levine, J. M. (1982). Socialization in small groups: Temporal changes in
individual-group relations. In L. Berkowitz (Ed.), Advances in experimental social psychology
(Vol. 15, pp. 137–192). New York: Academic Press.
Moreland, R. L., & Levine, J. M. (2001). Socialization in organizations and workgroups. In
M. E. Turner (Ed.), Groups at work: Theory and research (pp. 69–112). Hillsdale, NJ: Erlbaum.
Organ, D. W. (1988). Organizational citizenship behavior: The good soldier syndrome.
Lexington, MA: Lexington Books.
Organ, D. W., & Ryan, K. (1995). A meta-analytic review of attitudinal and dispositional predictors
of organizational citizenship behavior. Personnel Psychology, 48, 775–802.
Pratt, M. G. (1998). To be or not to be? Central questions in organizational identification. In
D. A. Whetten & P. C. Godfrey (Eds.), Identity in organizations: Building theory through
conversations (pp. 171–207). Thousand Oaks, CA: Sage.
Preacher, K. J., & Leonardelli, G. J (2003). Calculation for the Sobel test. An interactive calculation
tool for mediation tests. Retrieved from the Internet on 29 April 2005: preacher/sobel/so-
bel.htm" . http://www.unc.edu/ , preacher/sobel/sobel.htm
Reicher, S. D., Haslam, S. A., & Hopkins, N. (2005). Social identity and the dynamics of leadership:
Categorization, entrepreneurship and power in the transformation of social reality. Leadership
Quarterly, 16, 547–568.
Riketta, M. (2005). Organizational identification: A meta-analysis. Journal of Vocational Behavior,
66, 358–384.
Riketta, M., & Van Dick, R. (2005). Foci of attachment in organizations: A meta-analysis
comparison of the strength and correlates of work-group versus organizational commitment
and identification. Journal of Vocational Behavior, 67, 490–510.
Schneider, B. (1987). The people make the place. Personnel Psychology, 40, 437–453.
Shamir, B., R, J., & Arthur, M. B. (1993). The motivational effects of charismatic leadership: A self-
concept based theory. Organization Science, 4, 577–594.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and non-experimental studies:
New procedures and recommendations. Psychological Methods, 7, 422–445.
Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural equations models. In
S. Leinhart (Ed.), Sociological methodology (pp. 290–312). San Francisco: Jossey-Bass.
Ulber, D. (2001). Survey-Feedback-Instrument ODAS zur Organizationsdiagnose an Schulen –
Testgüte und explorative Analysen. In Ch. Hanckel, B. Jötten, & K. Seifried (Eds.), Schule
zwischen Realität und Vision. [Survey instrument for organizational diagnosis in schools –
psychometric properties and explorative analyses] (pp. 96–102). Bonn: Deutscher Psycholo-
gen Verlag.
Van Dick, R. (2004). My job is my castle: Identification in organizational contexts. In C. L. Cooper
& I. T. Robertson (Eds.), International review of industrial and organizational psychology
(Vol. 19, pp. 171–204). Chichester: Wiley.
Van Dick, R., Wagner, U., Stellmacher, J., & Christ, O. (2004). The utility of a broader
conceptualization of organizational identification: Which aspects really matter? Journal of
Occupational and Organizational Psychology, 77, 171–191.
Van Dick, R., Wagner, U., Stellmacher, J., & Christ, O. (2005). Category salience and its effects on
organizational identification. Journal of Occupational and Organizational Psychology, 78,
273–285.
150 Rolf Van Dick et al.
Van Knippenberg, D. (2000). Work motivation and performance: A social identity perspective.
Applied Psychology: An International Review, 49, 357–371.
Van Knippenberg, D., & Hogg, M. A. (2003). A social identity model of leadership effectiveness in
organizations. In B. Staw & R. M. Kramer (Eds.), Research in organizational behavior (Vol.
25, pp. 245–297). Greenwich, CT: JAI Press.
Van Knippenberg, D., & Van Knippenberg, B. (2005). Leader self-sacrifice and leadership
effectiveness: The moderating role of leader prototypicality. Journal of Applied Psychology,
90, 25–37.
Van Knippenberg, D., Van Knippenberg, B., De Cremer, D., & Hogg, M. A. (2004). Leadership, self,
and identity: A review and research agenda. Leadership Quarterly, 15, 825–856.