Full text of DOI:10.1348/014466600164480

British Journal of Social Psychology (2000), 39, 301±308 Printed in Great Britain
301
#
2000 The British Psychological Society
Leadership behaviour similarity as a basis of
selection into a management team
Geoå Eagleson,* Robert Waldersee and Ro Simmons
Centre for Corporate Change, Australian Graduate School of Management,
University of New South Wales, Australia
There are contexts where, in order to be eåective, a management team needs to
display diåerent leadership behaviours as well as possessing a wide range of
experiences and skills. But achieving heterogeneity of behaviours is di¬cult if there
are behavioural±similarity biases aåecting a selection process.
A sample of 126 participants completed a simulation of a selection process for a
management team. They compared selection criteria and chose either a task or
relations-oriented applicant for an advertised position as well as reporting their
perception of their own preferred leadership style. The results of the simulation
indicate that leadership behaviour can be a basis for similarity±attraction in the
selection of a management team.
An issue of concern to practising managers is how to select more eåective
management teams. One contributor to successful management team performance is
the heterogeneity of experience and skills within the team. Heterogeneity of
demographics and functional backgroundsin the top management team is important
for eåective decision-making (Hambrick, 1994; Jackson, Stone, & Alvarez, 1993) as
it enhances the search for information and provides the momentum for change
(Dutton & Duncan, 1987). The bene®ts of management team diversity ultimately
aåect bottom-line performance (Murray, 1989).
Drawing on Byrne’s (1971) similarity±attraction theory, both the Attraction±
Selection±Attrition Model (ASA; Schneider, 1987) and the Organizational De-
mography Model (Pfeåer, 1983) argue that individuals are attracted to, and more
likely to select, similar others. Speci®cally, the ASA model posits that attraction to
others is based on similarity in personality, attitudes and values. The Demography
model posits age, tenure, sex, race, educational levels, socio-economic background
and religion are the key variables on which similarity±attraction occurs (Jackson et
al., 1991; Tsui & O’Reilly, 1992).
A number of studies has demonstrated that hiring practices are biased by
demographic similarity (e.g. Graves & Powell, 1995; Westphal & Zajac, 1995).
Consistent with these ®ndings, top-level executives in large organizations have been
found to be of a similar age, have similar backgrounds, degrees and experience and
* Requests for reprints should be addressed to Professor Geoå Eagleson, Australian Graduate School of
!
Management, University of New South Wales, Australia 2052 (e-mail: G.Eagleson unsw.edu.au).

302
Geoå Eagleson et al.
do not represent the broader population of employees (e.g. Jackson et al., 1991;
Useem & Karabel, 1986; Westphal & Zajac, 1995).
Apart from experience and skill heterogeneity, there is some evidence that
heterogeneity of leadership behaviour style is also related to team performance. This
was initially reported in small group research where two distinct leaders have been
consistently identi®ed: one task-oriented and the other relations-oriented (Benne &
Sheats, 1948; Borgatta, Couch, & Bales, 1954; Fleishman, 1967). The task-oriented
leader organizes and structures the group to accomplish its task e¬ ciently by setting
up chains of communication and specifying working methods and standards. The
relations-oriented leader is concerned with maintaining harmonious, co-operative
relationships among team members who must work together to accomplish the task.
Groups where both leadership behaviours are present are likely to be more
productive and more cohesive (Bass, 1990).
There is also evidence that heterogeneity of leadership behaviour within a top
management team is related to organizational performance. When the internal
organizational environment is turbulent, such as during major change, task-oriented
leadership is needed to focus on results while relations-oriented leadership is needed
to facilitate the process (Nadler & Tushman, 1990). In a study of the introduction
of a quality programme within a hotel chain, diversity of leadership behaviours
across the steering committee of the change programme was associated with positive
outcomes (Waldersee, Simmons, & Eagleson, 1995). It would appear there are
bene®ts in consciously forming a management team to encompass a wide variety of
experiences, skills and leadership styles.
However, diversity of leadership behaviour styles may not be the natural state of
management teams. It could be that similarity±attraction occurs on the basis of
leadership behaviour. In many internal selection situations such as work team
formation, management coalitions, internal promotions or redeployment, the
leadership style of applicants is as well known as their attitudes, values and
demographics. If similarity±attraction occurs on the basis of leadership behaviour,
then the natural tendency of management teams would be toward homogeneity of
task or relations leadership style.
Importantly, similarity±attraction on the basis of behaviour is unlikely to be
related to similarity±attraction on other factors. Sheridan and Vredenburgh (1978)
found almost no relationship between task or relations behaviours of the leader and
leader characteristics or demographics, including age, experience, locus of control,
social relations with subordinates or status diåerence from subordinates. The degree
of demographic similarity is unlikely to be related to leadership behaviour similarity.
The current study had individuals complete a simulation of a selection process for
a management team. The aim of the study was to examine whether a similarity±
attraction eåect occurs on the basis of leadership behaviour as speci®ed in the
following hypothesis:
H1: Selectors are more likely to choose an applicant with the same leadership
style as themselves.
Although leader behaviour similarity may play a role in selection, numerous
cognitive processes of top team members, such as information processing and

Leadership behaviour similarity
303
cognitive categorization of leadership characteristics, are also operating (Bass, 1990;
Lord, 1985). The typical formal selection process attempts to make these processes
explicit through the use of selection criteria (Gerstein & Reisman, 1983). These
criteria should then drive the selection decision:
H2: Selectors are more likely to choose an applicant with a leadership style
similar to the selection criteria they have identi®ed as more important.
However, there is no indication in the literature about whether rating selection
criteria is likely to mediate or act independently of any similarity±attraction bias.
Since self-perceived leadership style is known to be stable over time (Greene, 1975),
one could argue that it could bias both the judgment of the selection criteria and the
applicant choice, so the authors here hypothesize:
H3: Selection bias due to behavioural similarity will be mediated through an
assessment of selection criteria.
Methods
A selection exercise was designed, based on common selection practices. Participants took the role of
General Manager of a courier company. They were told that a service quality programme was to be
introduced into the company. Therefore, the company had decided to advertise for a Service Quality
O¬cer. It was noted that the successful applicant would have responsibility for implementing the
programme jointly with the General Manager. The Service Quality O¬cer and General Manager would
thus form an informal executive.
An advertisement was presented which used the layout, criteria and wording typical of a sample of
newspaper advertisements for similar positions. Participants were ®rst asked to rate a list of selection
criteria on importance to the advertised position and then choose one of two applicants on the basis of
a written assessment, modelled on documentation provided by recruitment ®rms. They then completed
a Leadership Behaviour Description Questionnaire (LBDQ-XII; Stogdill, 1963) on themselves to
provide their own perception of their leadership style.
Participants
The participants were school alumni, graduating prior to 1990 (N 280) and with `Manager’ or
5
`Director’ in their job title, together with Masters of Business Administration (MBA) students (N
5
70). Of the participants, 80% were male. Ninety-®ve alumni (a 34% response rate) and 70 MBA
students (a 60% response rate) returned the forms. Thirty-eight MBA and 88 alumni returns were
usable, providing a sample of 126. The exercise was administered directly to MBA students and mailed
to alumni, together with a covering letter explaining details and process.
Measures
Selection criteria. Participants rated 14 selection criteria on their importance for the advertised position.
Seven of the attributes were relations-oriented and seven were task-oriented. Examples of relations
criteria included `people skills ’, `ability to resolve con¯ict’ and `supportive of staå’. Task criteria
included `results orientation’, `provides clear direction’ and `ability to set clear goals’. The selection
criteria were rated on a 5-point Likert scale (`very unimportant’ to `very important’).
Applicant pro®les. Two applicant pro®les were presented in random order. Demographic information
was included to simulate the selection process but all demographic variables were matched across the

304
Geoå Eagleson et al.
two pro®les. The task and relations behavioural orientation of the applicants was distinguished within
referee reports, using items from the LBDQ, as well as behavioural descriptors taken from Bass (1990).
The relations-oriented applicant’s report included:
`Mr. Roberts¼ treated all staå as his equal and promoted mutual trust¼ He consulted workers
on important matters and constantly put suggestions made by the group into operation¼’
The task-oriented applicant’s report included:
`Mr. Cooper¼made clear decisions about what work needed to be done and prioritized and
scheduled work accordingly¼He encouraged the use of uniform procedures¼He de®ned and
structured employees’ roles.’
Participant leadership style. The LBDQ-XII was used to identify the participant’s own perceived
leadership style. It provides scores on initiating structure (task-orientation) and consideration (relations-
orientation). Participants indicated the frequency with which they believed they engaged in the stated
behaviour on a 5-point Likert scale (`always’ to `never’). This instrument has good construct validity
(Stogdill, Goode, & Day, 1962), content validity (Fleishman, 1953) and high test±retest reliability
(Green, 1975). The LBDQ has 10 items relating to task behaviour and 10 relating to relations behaviour.
Manipulation check. The task and relations orientation of the applicants was rated by each participant on
a 5-point scale (`very low’ to `very high’).
Demographics. Not surprisingly, given the origin of the two groups of participants (MBA students and
alumni who graduated before 1990), there was evidence of diåerences between the two groups in age,
management experience and selection experience.
Results
.
Participants judged themselves to be more relations- than task-oriented (t(125) 3 3,
5
001). They also perceived that the job required more relations- than task-
!
.
p
!
.
.
oriented behaviour (t(125) 10 3, p 001). Overall, 77% of participants chose the
5
relations-oriented candidate. The alumni scored themselves higher on task
!
signi®cant sex eåect or interaction. Females scored themselves higher than males on
.
.
behaviours than the MBAs (F(1, 122) 6 7; p
01), though there was no
5
!
.
.
05) with no signi®cant group eåect or
relations behaviours (F(1, 122) 6 2; p
interaction.
5
Manipulation assessment
The task applicant was rated higher on the task behaviour scale than the relations
!
.
.
001). The relations applicant was rated higher on the
applicant (t(125) 10 92, p
5
!
.
.
001).
relations behaviour scale than the task applicant (t(125) 21 66, p
5
Association between participant leader style, criteria and applicant choice
Because the behavioural style of the applicants will be less important when the
General Manager is able to perform both task and relations behaviours, `balanced ’
participants were removed. Participants were categorized as `balanced ’ if their task
and relations scores diåered by less than 5%. Though somewhat arbitrary, a
sensitivity analysis showed that the results obtained were not signi®cantly diåerent
when other cut-oå points were used. Eighty-nine participants were left, 32 of them
MBAs and 57 alumni. Summary statistics for the reduced data set are given in
Table 1.

Leadership behaviour similarity
305
89)
Table 1. Summary statistics when `balanced ’ participants were removed (N
5
Correlations
M
SD
1
2
3
4
1. Self-assessed task
behaviour
37.28
4.35
Ð
2. Self-assessed relations
behaviour
39.35
3.41
.20
Ð
Ð
2
3. Task criteria
4. Relations criteria
29.58
32.78
3.29
2.06
.31**
.04
.07
.24*
Ð
.11
2
Ð
!
!
* p .05; ** p 0.1.
Table 2. Logit regression for choice of relations-oriented applicant
Applicant choice
regressed on:
Change in
deviance
a
#
Coe¬cient
4.88
SE
d.f.
R
Self-assessed behavioural scores
Intercept
Behaviour task score
Behavior relations score
Criteria scores
4.29
.07
.08
.21**
.11
2
15.1***
13.4**
2
2
.16
.13
Intercept
4.00
4.68
.13
.12
Criteria task score
Criteria relations score
Behavioural and criteria scores
Intercept
.37**
.25*
2
6.85
6.10
.07
.09
.14
.14
Behaviour task score
Behaviour relations score
Criteria task score
Criteria relations score
.19**
.07
2
.31*
.25
2
23.3***
4
.24
!
!
!
* p 0.5; **p .01; ***p .001.
a
#
The squared residuals form of R for logistic regression was used (see Mittlbock & Schemper, 1996).
Logit regression was ®tted to examine the relationship between leadership style
!
Coe¬ cients were in the predicted direction, with task-orientation negatively
.
and applicant choice. Deviance reduction was signi®cant ( p
001; see Table 2).
!
relations-orientation positively associated with the choice of the relations-oriented
.
associated with the choice of the relations-orientation applicant (p
01), and
#
applicant. The direct analogue of the linear regression R was used: the proportional
reduction in the sum of squares when the predicted probabilities are used instead of
#
the sample average (Mittlbock & Schemper, 1996). The R was 16%, con®rming

306
Geoå Eagleson et al.
hypothesis H1. A second logit regresssionexamined the relationship between criteria
scores and applicant choice, con®rming hypothesis H2 (see Table 2).
Finally, both leadership and criteria scores were used to predict applicant choice
!
#
.
(see Table 2). The reduction in deviance was signi®cant (p
001), and R was 24%.
The coe¬ cients were in the predicted directions, and little changed from their
previous values. Despite some signi®cant correlations between the predictor
#
variables, both the change in deviance and R are close to being additive. It would
appear that the leadership and criteria scores could be in¯uencing choice
independently. Hypothesis H3 is not supported.
Discussion
That similarity±attraction occurs in a management context on the basis of
demographic and attitudinal variables has long been established. The results of the
current study indicate that similarity±attraction in the context of a management team
may also occur between those who exhibit similar leadership behaviours, limiting the
diversity of leader behaviours across the team. The results also suggestthat this eåect
is largely independent of the cognitive processes and biases which occur in the
identi®cation of selection criteria and the eåect of criteria assessment on the selection
decision. With a tendency away from complementarity of behaviours, de®ciencies in
one member will not be oåset by strengths in another.
However, diversity of leadership behaviours in a management team may not
always be desirable. Diverse teams often have low consensus, low aåective
acceptance of decisions (Amason, 1996), high integration costs and high turnover
(Milliken & Martins, 1996). Furthermore, some leadership theories (e.g. Fiedler,
1967) argue that most of the time only one of the task or relations styles is necessary,
contingent on the situation. In such situations, teams with diverse leadership styles
may be less eåective than those with a single, consistent style.
When heterogeneity is desirable, similarity±attraction is only one of the processes
acting on selection within teams. Other processes may sway selection toward
diversity rather than similarity. For example, there is a tendency to select people with
values that match the organizational environment (Chatman, 1991). When the
organizational situation changes, selectors may view the ®t between the management
team leadership style and the organization as low. In this context, applicants diåerent
in behavioural style from the current management team, but ®tting the needs of the
new organization, may be selected.
Diversity may also increase when the position and power of an individual leader
is of paramount concern. Charismatic leaders, for example, use their strong social
capabilities (Lord & Hall, 1992) to manage their impression on followers (Gardner
& Avolio, 1998). Additional charismatic leaders in a management team would split
and dilute this process. Here, selection could be expected to be based on contribution
to instrumental leadership, cementing rather than challenging the power of the
primary leader (Nadler & Tushman, 1990).
Overall, diversity in a management team is intrinsically neither good nor bad.
Rather, the need for diversity is contingent upon the context. But where
heterogeneity of leadership behaviour is important, behavioural similarity±attraction

Leadership behaviour similarity
307
may negatively aåect the formation of eåective management teams. If managers
judge diversity to be important, they need to be cognisant of the forces and biases
against diversity, including behavioural diversity, and to respond accordingly.
Acknowledgements
The authors wish to acknowledge the help of the Editor and referees whose comments and
suggestions greatly improved the article. They also acknowledge the support of the Australian Research
Council through its funding of the Centre for Corporate Change.
A lengthier version of this study is available from the authors on request.
References
Amason, A. C. (1996). Distinguishing the eåects of functional and dysfunctional con¯ict on strategic
decision making: Resolving a paradox for top management teams. Academy of Management Journal, 39,
123±148.
Bass, B. M. (1990). Bass and Stogdill’s handbook of leadership : Theory, research and managerial applications (3rd
ed.). New York: Free Press.
Benne, K. D., & Sheats, P. (1948). Functional roles of group members. Journal of Social Issues, 2, 42±47.
Borgatta, E. F., Couch, A. S., & Bales, R. F. (1954). Some ®ndings relevant to the great man theory
of leadership. American Sociological Review, 19, 755±759.
Byrne, D. (1971). The attraction paradigm. New York: Academic Press.
Chatman, J. (1991). Matching people and organisations: Selection and socializations in public
accounting ®rms. Administrative Science Quarterly, 36, 459±484.
Dutton, J. E., & Duncan, R. B. (1987). The creation of momentum for change through the process of
strategic issue diagnosis. Strategic Management Journal, 8, 279±295.
Fiedler, F. E. (1967). A theory of leadership eåectiveness. New York: McGraw-Hill.
Fleishman, E. A. (1953). Leadership climate, human relations training and supervisory behaviour.
Personnel Psychology, 6, 205±222.
Fleishman, E. A. (1967). Development of a behaviour taxonomy for describing human tasks: A
correlational-experimental approach. Journal of Applied Psychology, 51, 1±10.
Gardner, W., & Avolio, B. (1998). The charismatic relationship : A dramaturgical perspective. Academy
of Management Review, 23, 32±64.
Gerstein, M., & Reisman, H. (1983). Strategic selection : Matching executives to business conditions.
Sloan Management Review (Winter), 33±49.
Graves, L. M., & Powell, G. N. (1995). The eåect of sex similarity on recruiters’ evaluations of actual
applicants: A test the similarity±attraction paradigm. Personnel Psychology, 48, 85±98.
Greene, C. N. (1975). A longitudinal analysis of relationships among leader behaviour and subordinate
performance and satisfaction. Academy of Management Proceedings, 433±440.
Hambrick, D. C. (1994). Top management groups: A conceptual integration and reconsideration of the
`team ’ label. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behaviour (Vol. 16, pp.
171±214). London: JAI Press.
Jackson, S. E., Brett, J. F., Sessa, V. I., Cooper, D. M., Julin, J. A., & Peyronnin, K. (1991). Some
diåerences make a diåerence: Individual dissimilarity and group heterogeneity as correlates of
recruitment, promotions and turnover. Journal of Applied Psychology, 76, 675±689.
Jackson, S. E., Stone, V. K., & Alvarez, E. B. (1993). Socialization amidst diversity: The impact of
demographics on work team old-timers and newcomers. In L. L. Cummings & B. M. Staw (Eds.),
Research in organizational behaviour (Vol. 15, pp. 45±109). London: JAI Press.
Lord, R. G. (1985). An information processing approach to social perceptions, leadership perceptions
and behavioural measurement in organizational settings. In L. L. Cummings & B. M. Staw (Eds.),
Research in organizational behaviour (Vol. 7, pp. 87±128). London: JAI Press.
Lord, R. G., & Hall, R. J. (1992). Contemporary views of leadership and individual diåerences.
Leadership Quarterly, 3, 137±157.

308
Geoå Eagleson et al.
Milliken, F., & Martins, L. (1996). Searching for common threads: Understanding the multiple eåects
of diversity in organizational groups. Academy of Management Review, 21, 402±433.
Mittlbock, M., & Schemper, M. (1996). Explained variation for logistic regression. Statistics in Medicine,
15, 1987±1997.
Murray, A. (1989). Top management group heterogeneity and ®rm performance. Strategic Management
Journal, 10, 125±141.
Nadler, D., & Tushman, M. L. (1990). Beyond the charismatic leader: Leadership and organizational
change. California Management Review, 32, 77±97.
Pfeåer, J. (1983). Organizational demography. In L. L. Cummings & B. M. Staw (Eds.), Research in
organizational behaviour (Vol. 5, pp. 299±357). London: JAI Press.
Schneider, B. (1987). The people make the place. Personnel Psychology, 40, 437±453.
Sheridan, J., & Vredenburgh, D. (1978). Predicting leadership behaviour in a hospital organisation.
Academy of Management Journal, 21, 679±689.
Stogdill, R. M. (1963). Manual for leadership behaviour description, questionnaireÐForm XII: An experimental
revision. Columbus, OH: Bureau of Business Research, College of Commerce and Administration,
The Ohio State University.
Stogdill, R. M., Goode, O. S., & Day, D. R. (1962). New leader behaviour, description subscale.
Journal of Psychology, 56, 259±269.
Tsui, A. S., & O’Reilly, C. A. (1992). Being diåerent: Relational demography and organizational
attachment. Administrative Science Quarterly, 37, 549±579.
Useem, M., & Karabel, J. (1986). Pathways to corporate management. American Sociological Review, 51,
184±200.
Waldersee, R., Simmons, R., & Eagleson, G. (1995). Pluralistic leadership in service change programs:
Some preliminary ®ndings. In D. P. Moore (Ed.), Academy of Management Best Papers Proceedings (pp.
296±300). Vancouver: Omnipress.
}
Westphal, J. D., & Zajac, E. J. (1995). Who shall govern? CEO Board power, demographic similarity,
and new director selection. Administrative Science Quarterly, 40, 60±83.
Received 10 September 1997; revised version received 14 May 1999