See

discussions, stats, and author profiles for this publication at:

https://www.researchgate.net/publication/232471264

The role of social and communication

skills in leader emergence and
effectiveness

Article in Group Dynamics Theory Research and Practice · May 2003

Impact Factor: 0.88 · DOI: 10.1037/1089-2699.7.2.83

CITATIONS READS

65 1,226

4 authors, including:

Ronald E. Riggio
Claremont McKenna College
111 PUBLICATIONS 3,423 CITATIONS

SEE PROFILE

Charles Salinas
University of Alaska Southeast
5 PUBLICATIONS 148 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Ronald E. Riggio
letting you access and read them immediately. Retrieved on: 10 June 2016
Group Dynamics: Theory, Research, and Practice Copyright 2003 by the Educational Publishing Foundation
2003, Vol. 7, No. 2, 83–103 1089-2699/03/$12.00 DOI: 10.1037/1089-2699.7.2.83

The Role of Social and Emotional Communication Skills in Leader

Emergence and Effectiveness

Ronald E. Riggio Heidi R. Riggio

Claremont McKenna College Pomona College

Charles Salinas Emmet J. Cole

California State University, Long Beach Orange County Fire Department

Three studies explored the role that basic social/communication skills play in leader
emergence and effectiveness. In Study 1, 218 undergraduate students were adminis-
tered self-report measures of social/communication skills and extraversion, worked in
small groups on a problem-solving task, and elected leaders at task midpoint. Coders
measured leaders’ verbal communication. Groups elected leaders who spoke most and
were extraverted but were not more socially skilled. In Study 2, leaders were selected
on the basis of their possession of communication skills and led small groups in 2 tasks.
More skilled communicators were rated as more effective leaders, but they did not lead
more productive groups. Study 3 examined fire service leaders. Social skills were
related to satisfaction with the leader but related to leader performance only for higher
level leaders.

Since the inception of scientific research on Sometimes, possession of communication

leadership, skill in interpersonal communication skills is inferred from certain communication-
has been mentioned as a key element both in related personality traits, such as sociability,
predicting who emerges as a leader and in de- extraversion, nurturance, or assertiveness. A
termining the effectiveness of leaders. For in- more detailed (and perhaps more accurate) mea-
stance, Stogdill’s (1974) review of early re- surement of leader communicative abilities is
search on leadership emphasized the impor- made by having trained professionals judge in-
tance of communication skill in both leader dividuals’ communication competence. This
emergence and leader effectiveness. Bass method is used routinely in assessment centers
(1990) and Kanter (1983) both noted apparent (Bray, Campbell, & Grant, 1974; Finkle, 1976;
connections between communication skill/com- Howard & Bray, 1990). However, a major prob-
petence and leader and managerial effective- lem with these judged assessments of commu-
ness. However, a major difficulty in studying nication abilities is that they focus almost ex-
the communication skills of emergent and es- clusively on oral communication skills, partic-
tablished leaders has been the operationaliza- ularly speaking ability. Decoding skills,
tion of communication skills or communication sensitivity to others’ communication, and the
competence. role of nonverbal skills are almost totally over-
looked (Hollander, 1985). An additional prob-
lem with these judgments or ratings of leader
Ronald E. Riggio, Kravis Leadership Institute, Claremont communication abilities is that the assessments
McKenna College; Heidi R. Riggio, Department of Psy- are usually made in specific situations. Individ-
chology, Pomona College; Charles Salinas, Department of
Psychology, California State University, Long Beach; Em- uals who appear to be effective communicators
met J. Cole, Orange County Fire Department, Orange in one situation may not be effective in other
County, California. situations. For example, an individual may pos-
Correspondence concerning this article should be ad- sess a great deal of task-relevant knowledge in
dressed to Ronald E. Riggio, Kravis Leadership Institute,
Claremont McKenna College, 850 Columbia Avenue,
one situation that stimulates greater communi-
Claremont, California 91711. E-mail: ronald.riggio@ cation of this knowledge. However, in another
claremontmckenna.edu situation, where task-related knowledge is
83
84 RIGGIO, RIGGIO, SALINAS, AND COLE

lower, this same individual may communicate gence as would be possession of tactful com-
relatively infrequently. Finally, observational munication skills— being able to say the right
assessments of communication skill are usually things at the right time. In fact, research on the
difficult, time consuming, and expensive to role of self-monitoring has found that high self-
obtain. monitors are more likely to emerge as leaders
The purpose of the present research was to than those low on self-monitoring (Dobbins,
explore the role that possession of basic com- Long, Dedrick, & Clemons, 1990; Ellis, 1988;
munication/social skills plays in both the emer- Ellis, Adamson, Deszca, & Cawsey, 1988).
gence of leaders in novel situations and the Self-monitoring, according to Snyder (1974,
effectiveness of individuals in leadership roles. 1987), is the ability to “read” social situations
By using a multidimensional model of basic and to alter one’s own behavior to fit in and act
communication skills (Riggio, 1986, 1989), we appropriately in social situations. Presumably,
hoped to gain insight into both how possession high self-monitors can adapt to situational con-
of communication skills/competence assists an straints, and it is this ability that enables them to
individual in attaining a leadership position and emerge as leaders. Importantly, research has
the connection between communication skills shown strong connections between self-moni-
and leader effectiveness. toring and possession of basic communication
skills (Riggio, 1986; Riggio & Friedman, 1982).
Communication Skills and Leader It is anticipated that exceptionally skilled com-
Emergence municators would, like high self-monitors, be
more adaptable to situational constraints and
The majority of early studies that attempted thus more likely to emerge as leaders than
to find predictors of leader emergence in groups would persons lacking basic communication
focused on personality traits, and they met with skills.
little success (Bass, 1990; Mann, 1959). Such Research in personnel selection has also con-
trait approaches to leadership were criticized sistently found connections between the posses-
for their inability to account for situational fac- sion of communication skills and hiring, albeit
tors that could also influence leader emergence most of this research has focused on evaluations
(Kenny & Zaccaro, 1983). Another possibility made in hiring interviews. For instance, nonver-
is that the trait characteristics studied in these bally expressive individuals are evaluated more
early investigations did not link up well with the favorably in hiring interviews than are nonex-
behaviors emitted by emergent leaders. For ex- pressive individuals (Gifford, Ng, & Wilkinson,
ample, it is a well-known and consistent finding 1985). In addition, persons possessing good
that emergent leaders simply communicate role-playing communication skills and appli-
more in group settings than do other members cants who possess impression management
of the group (Bass, 1949; Mullen, Salas, & skills also perform better in hiring interviews
Driskell, 1989; Wentworth & Anderson, 1984). than persons lacking these skills (Fletcher,
It seems likely that characteristics such as ex- 1989; Riggio & Throckmorton, 1988). If it is
traversion, sociability, and possession of basic assumed that some workplace leaders are se-
communication skills would be related to lected via traditional methods such as hiring
greater levels of communication in leaderless interviews, then there may be direct connections
discussion groups and thus would be predictive between communication skills and being se-
of leader emergence in these groups (see Lord, lected for certain leadership positions.
DeVader, & Alliger, 1986). However, in groups The first study in the present research ex-
that do not rely so heavily on quantity of com- plored the role of basic communication skills in
munication, sheer volume of communication predicting leader emergence in small discussion
might not be predictive of leader emergence. In groups. Also explored were a traditional person-
task-oriented groups, for example, quality of ality trait, extraversion, as well as connections
communication rather than quantity of commu- between quantity of communication, possession
nication might better predict leader emergence of communication skill, extraversion, and
(Sorrentino & Boutillier, 1975). It would be leader emergence. We expected, consistent with
expected, then, that mere extraversion or socia- previous research, that emergent leaders in
bility would not be as important in leader emer- novel groups would be those who communi-
 LEADER COMMUNICATION SKILLS 85

cated most frequently. However, we also ex- skills (also referred to as interpersonal sensitiv-
pected that persons who were skilled commu- ity) are crucial for the development of good
nicators, as defined by possession of a broad interpersonal relationships (Riggio, 2001; Rig-
range of communication competencies, would gio & Zimmerman, 1991). The importance of
be more frequently chosen as leaders by their good leader–subordinate relationships has been
group members, thus supporting the notion that shown to be linked not only to traditional in-
both quantity and quality of communication are dexes of leader effectiveness, such as work-
important in leader emergence. group productivity, but also to subordinates’
satisfaction with the leader (Kahn & Katz,
Communication Skills and Leader 1952). Indeed, entire theories of leadership,
Effectiveness such as the leader–member exchange model
(Dansereau, Graen, & Haga, 1975; Graen, No-
Another important issue in leadership is the vak, & Sommerkamp, 1982), are predicated on
role of communication skills in leader effective- the notion that the quality of leader–subordinate
ness. Although leadership theorists consistently relationships is central to leader effectiveness.
emphasize the importance of the leader’s ability Furthermore, although research on charismatic
to communicate both verbally and nonverbally, and transformational leadership shows signifi-
there has been very little research directly ex- cant effects of leader charisma on group task
amining individual differences in leader com- performance (e.g., Howell & Frost, 1989;
munication skills or exploring how a leader’s Waldman, Bass, & Yammarino, 1990), charis-
possession of communication skills causes a matic–transformational leaders may have an
leader to be effective. This is particularly true even stronger effect on group members’ satis-
for the nonverbal aspects of communication. faction with and evaluation of the leader (e.g.,
For example, a review of research on the role of Lowe, Kroeck, & Sivasubramaniam, 1996),
nonverbal communication in leadership and
presumably by virtue of these leaders’ abilities
management yielded no studies researching the
to communicate effectively to arouse, inspire,
connections between a leader’s nonverbal com-
and motivate followers.
munication style or ability and the leader’s ef-
In addition to encoding and decoding skills,
fectiveness, although a few studies have looked
at how nonverbal communication affects per- effective leaders must possess good role-play-
ceptions of leaders (P. J. DePaulo, 1992). We do ing skills. For instance, effective leaders need to
know, however, that being nonverbally, or emo- be skilled in self-presentation and impression
tionally, expressive is related to positive evalu- management (Chemers, 1997; Leary, 1989). In
ations in social situations (B. M. DePaulo & fact, one theory of leadership, the path– goal
Friedman, 1998; Friedman, Prince, Riggio, & theory, revolves around the notion that effective
DiMatteo, 1980; Riggio, 1986) and is associ- leaders must play particular roles to enable
ated with perceptions of an individual’s cha- group members to achieve desired goals (see
risma (Friedman, Riggio, & Casella, 1988; Rig- House, 1971; House & Mitchell, 1974). Such
gio, 1987, 1998; see also Gardner & Avolio, role-playing skills involve highly developed
1998). communication skills, similar to the self-moni-
There are a number of ways that communi- toring skills discussed earlier.
cation skills can influence leader effectiveness. The second study in the present series of
First, work by Mintzberg (1973) and others has investigations explored how possession of en-
emphasized that workplace leaders and manag- coding, decoding, and role-playing skills influ-
ers spend most of their time communicating to enced rated leader effectiveness in small
subordinates—providing information, instruct- groups, and how possession of these communi-
ing, directing, coordinating, and giving feed- cation skills affected group members’ percep-
back (Katz & Kahn, 1966). Clearly, skills in tions of their leader. This is the first investiga-
sending, or encoding, verbal messages are crit- tion to explicitly explore how possession of
ical for effective leaders. Also important for these basic communication skills relates to
effective workplace leadership are interpersonal leader effectiveness. A third study explored
skills that help build good relationships with how communication skills of actual work-group
subordinates and peers. Listening, or decoding, leaders (officers in the fire service) correlated
86 RIGGIO, RIGGIO, SALINAS, AND COLE

with subordinates’ ratings of their leaders’ be in balance. In other words, possessing high
performance. levels of one basic communication skill, say
emotional expressiveness, while lacking an-
other basic communication skill, emotional con-
A Model for Basic Communication Skills trol (the ability to regulate and control emotions
and emotional expressions), is a form of imbal-
A major difficulty in studying the communi- ance (see Riggio, 1989). The individual effec-
cation skills of emergent and established leaders tively expresses emotions but is unable to con-
has been the operationalization of communica- trol their expression. The individual is “emo-
tion skills or communication competence. Many tionally transparent”— others can always tell
studies have used personality instruments, such what the person is feeling, and the individual is
as measures of extraversion or sociability, to unable to control or stifle the expression of felt
infer possession of communication competen- emotions. Research has suggested that such
cies (e.g., Mumford et al., 2000; Zaccaro, Mum- communication/social skill imbalances are
ford, Connelly, Marks, & Gilbert, 2000). Other linked to ineffective social relationships and
studies have used global ratings of communica- poor psychosocial adjustment (Perez, Riggio, &
tion skill, such as ratings made of leaders in Kopelowicz, 2002; Riggio, Watring, & Throck-
assessment centers (e.g., Thornton, 1992). The morton, 1993; Riggio & Zimmerman, 1991). In
present investigation used a comprehensive ap- addition to total and subscale scores, the SSI
proach to the measurement of communication provides a score for communication/social skill
abilities, and one that assesses communication imbalance.
competence across situations. This approach Importantly, validity studies for the SSI show
views communication/social competence as that scores on the self-report SSI do indeed
composed of several basic communication correlate significantly with performance-based
skills. These skills are of three types: skill in measures of communications skill. For exam-
sending (encoding) messages, termed expressiv- ple, the Emotional Expressivity Scale of the SSI
ity; skill in receiving (decoding), labeled sensi- was correlated with both posed and spontaneous
tivity; and skill in regulating communication, or encoding of basic emotions (Tucker & Riggio,
control. These three basic skills operate in two 1988), and the Emotional Sensitivity Scale of
domains: the verbal, or social, domain, and the the SSI was significantly positively correlated
nonverbal, or emotional, domain (Riggio, with decoding skill as assessed by the Profile of
1986). These six basic communication skills— Nonverbal Sensitivity (Rosenthal, Hall, DiMat-
emotional expressivity, emotional sensitivity, teo, Rogers, & Archer, 1979). These studies and
emotional control, social expressivity, social others suggest that the SSI does indeed lead to
sensitivity, and social control— can be assessed a general assessment of general communication
through a self-report instrument, the Social skill/competence (see Riggio & Riggio, 2001;
Skills Inventory (SSI; Riggio, 1989; Riggio & see also Spitzberg & Cupach, 1989).
Carney, 2003).
The total score on all six scales of the SSI
represents a measure of global communication/ Communication Skills and Social and
social skill or competence. In a series of inves- Emotional Intelligence
tigations, total score on the SSI has been theo-
retically and empirically linked to variables Early investigations of characteristics of
such as empathy, self-esteem, and social insight emergent and effective leaders focused on com-
(Riggio, 1986; Riggio, Messamer, & Throck- munication/social skills, more traitlike con-
morton, 1991; Riggio, Throckmorton, & DePa- structs such as sociability and extraversion, and
ola, 1990; Riggio, Tucker, & Coffaro, 1989)— social intelligence (see Bass, 1990, for an over-
variables that are themselves associated with view). More recently, there has been a resur-
leadership (Bass, 1990). gence of interest in social intelligence and lead-
Although possessing high levels of basic ership (e.g., Zaccaro, 2002; Zaccaro, Gilbert,
communication skills is important for persons Thor, & Mumford, 1991). A newer, related con-
to be communicatively competent, it is also struct is emotional intelligence. Both social and
important that these basic communication skills emotional intelligence are closely intertwined
 LEADER COMMUNICATION SKILLS 87

with notions of interpersonal communication A Resurgence of Interest in Trait

competence. From the leadership perspective, Approaches to Leadership
social and emotional intelligence represent the
“people skills” possessed by effective leaders In many ways, there has been somewhat of a
(see Bass, 2002; Zaccaro, 2002). Caruso, resurgence in the trait approach to studying
Mayer, and Salovey (2002) suggested why leadership (Riggio, 2002). In 1983, Kenny and
emotional intelligence is critical to leader effec- Zaccaro suggested that stable personal charac-
tiveness. They stated that leaders need to under- teristics accounted for an enormous amount of
stand their own emotions and the emotions of the variance in leader emergence. However, the
followers in order to encourage and motivate traits investigated by these researchers are more
others and to encourage follower creativity. In complex and multifaceted than the traits studied
by early leadership scholars. These modern
addition, the quality of leader and follower re-
traits involve complex social abilities such as
lationships is dependent on leader emotional
self-monitoring (Zaccaro, Foti, & Kenny,
intelligence (see Caruso et al., 2002; Goleman, 1991), behavioral flexibility (Kenny & Zaccaro,
1998a, 1998b). Moreover, Murphy (2002) ar- 1983), interpersonal acumen (Aditya & House,
gued that social and emotional intelligence, as 2002), emotional and social intelligence, and
represented by measures such as the SSI, helps communication competence.
in leader self-regulation— critical for leader ef- The present studies represent an initial at-
fectiveness in stressful situations. tempt to directly examine the relationship be-
Riggio’s communication/social skill model tween possession of basic communication
(Riggio & Carney, 2003), in some ways, paral- skills/competence and leader emergence and ef-
lels the concepts of emotional and social intel- fectiveness. Using the SSI, we were able to
ligence. Three of the SSI subscales deal with explore the role of communication competence
skill in nonverbal and emotional communica- in predicting leader emergence and leader ef-
tion (emotional expressivity, emotional sensi- fectiveness in much the same way that leader
tivity, and emotional control). The Mayer and personality traits have been previously ex-
Salovey (1997) ability model of emotional in- plored. In addition, attempts were made to gain
telligence includes the ability to perceive and some insight into the communicative behaviors
understand emotions (similar to emotional sen- that mediate the relationship between posses-
sitivity), the ability to regulate emotions (simi- sion of communication skills and leader emer-
lar to emotional control), and the ability to gence and effectiveness.
generate emotions (similar to emotional expres-
sivity), among other emotional abilities. Clearly, Study 1
there is overlap in the two models, and research
is underway directly examining the relation- A first study was designed to explore the role
ships between the SSI and recent measures of that possession of basic communication skills
emotional intelligence. plays in leader emergence in small, interacting
groups. It was hypothesized that possession of
There is also conceptual overlap between the
basic communication skills, as measured by to-
SSI model and social intelligence. Zaccaro
tal score on the SSI, would be predictive of who
(2002) discussed various “social capacities” would be chosen as leader by other group mem-
that are components of social intelligence. bers. However, factors other than mere posses-
These include ability to “perceive and interpret sion of communication skill were also explored.
social cues” (p. 38; similar to the social sensi- As mentioned earlier, one of the most con-
tivity dimension of the SSI model) and social sistent predictors of leader emergence in new
interaction competencies. The social expressiv- groups is talkativeness. The more an individual
ity and social control dimensions of the SSI verbalizes in a group interaction, the more
model deal with ability to engage others in likely the individual is to emerge as a leader.
social interaction and social role-playing skill, This has been referred to as the “babble hypoth-
respectively, and represent some, but not all, of esis” (Bass, 1990, p. 93). Yet the links between
the social interaction competencies suggested communication skill, verbosity, and leader
by Zaccaro. emergence are unclear. It was expected that
88 RIGGIO, RIGGIO, SALINAS, AND COLE

skilled communicators would engage in more the second session, during which they were
relevant communication than would nonskilled informed that they would be participating in
persons, and it was anticipated that the verbal small groups and that the group would be vid-
communication of skilled individuals would be eotaped. For each session, 6 participants were
of higher quality than the verbal messages of signed up; however, if one participant did not
nonskilled communicators. Therefore, we ex- show, the experiment continued with 5 partici-
pected that more talkative individuals would be pants. If more than one participant did not show,
more likely to emerge as leaders than relatively the experiment was rescheduled or cancelled. A
silent persons would. We also anticipated that total of 39 usable groups participated.
the communication of socially skilled individu- During the second session, participants were
als would be of higher quality than that of greeted by the experimenter, who affixed a large
nonskilled persons, and that a combination of nametag with participants’ first names to their
communication skill and talkativeness would be shirts and instructed them to take any seat. Six
the best predictor of leader emergence. There- chairs were arranged in a semicircle facing a
fore, this study also examined the relationships video camera that was out of immediate site in
between possession of communication skills in an adjoining room. Participants were then each
both leaders and nonleaders, and the quantity given a version of a training exercise problem
and quality of verbal exchange in discussion developed by the U.S. Air Force. Participants
groups. were presented with the scenario of being a
passenger on an airplane that has just crash-
Method landed in the desert. A list of 15 items (e.g.,
water, parachute, compass) was presented, and
Participants. Participants were 218 under- participants were asked to independently rank-
graduate students (64 men and 154 women) order the items in terms of their survival value.
recruited from introductory psychology courses
This problem or variations of it have been used
who participated in exchange for course credit.
in numerous studies of group interaction and
Measures. In initial sessions, large groups
leader emergence. Participants were given 10
of students (approximately 30 –50) completed
min to work on this problem and were in-
the original version of the Social Skills Inven-
tory, using a 9-point response scale (Riggio, structed not to talk to one another during this
1986) and the Eysenck Personality Inventory period. The experimenter then told the partici-
(EPI) Extraversion Scale (Eysenck & Eysenck, pants that they would spend the next 10 min
1968). discussing the problem. The video camera re-
The Social Skills Inventory is a 90-item, self- corded this 10-min session. At the end of the
report instrument that provides a multidimen- group discussion, participants were told that
sional assessment of global social skills. For they would choose a group leader to continue
this initial study, the total SSI score was used, the discussion and to help the group develop a
with the original 9-point response scale, ranging group ranking of the survival items. The group
from not at all true of me to very true of me. leader would be responsible for leading the
This total score on the SSI reflects global social group, recording the final group ranking, and
skill or competence (Riggio et al., 1991). Inter- presenting the ranking in an individual video-
nal consistency of the total SSI was quite high taped presentation. Participants were then told
(␣ ⫽ .97). to vote for their top three choices for group
The EPI is a 90-item, self-report instrument leader by secret, written ballot. The experi-
that measures personality from Eysenck’s tridi- menter tallied the votes (giving 3 points for a
mensional model of Extraversion, Neuroticism, first choice, 2 points for a second choice, and 1
and Psychoticism. Participants respond to the point for a third choice) and announced the
items by answering yes or no. Only the 21 items participant who received the highest vote total.
from the Extraversion scale were used in the Each member thus received a “leader potential
present study. Internal consistency of this scale score” based on the total tally of group mem-
was satisfactory (KR20 ⫽ .71). bers’ votes for each participant. This process
Procedures. Upon completion of the self- was a simplified version of a procedure used by
report measures, participants were signed up for Bass (1949) to have participants “vote” for the
 LEADER COMMUNICATION SKILLS 89

group member who seemed to contribute the correct total. The last category of verbal cues,
most to the leaderless group discussion. task-irrelevant statements, occurred very infre-
The group then continued discussion for an quently and thus was dropped from analyses. In
additional 10 min with the selected participant addition, there was an extremely high correla-
playing the role of group leader. At the end of tion (r ⫽ .88) between amount of speaking and
this discussion, the group leader was taken to an speaking turns. Because amount of speaking
adjoining room for the videotaped presentation was the more precise measure, speaking turns
of the group’s ranking (which was used for was not included in the analyses. It is important
another study). Participants were told the pur- to note that of the four remaining categories of
poses of the experiment and were dismissed. coded statements, directive statements were by
The videotapes of the first 10 min of each far the most frequently coded verbal interac-
group’s discussion (before a leader was se- tions, constituting about 80% of the leader’s
lected) were transcribed to obtain the verbal verbal communication.
interchange, and the transcriptions were coded
to determine which participant was speaking
and to whom the message was being sent. Results
From the videotapes and the transcription, the
following variables were extracted: It was immediately apparent that there was a
gender bias in whom the groups chose as lead-
Amount of speaking: the sheer number of ers. By virtue of both the university’s popula-
words spoken by each participant tion (nearly 60% female) and the composition
of introductory psychology courses (about 70%
Speaking turns: the number of times each female), more women than men participated in
participant spoke the experiment. However, keeping with tradi-
tional views of the “leadership role,” males
Directive statements: the number of times were chosen disproportionately more often as
a participant made statements of fact or group leaders (19 male leaders vs. 20 female
opinion, gave instructions, or summarized leaders), ␹2(1, N ⫽ 1) ⫽ 8.6, p ⬍ .01. Despite
discussion (i.e., any task-related statements) this obvious bias, we anticipated that the key
Questions asked: the number of questions variables of extraversion and possession of
asked of other group members (e.g., “What communication skills, as represented by total
do you think?”; “What should we do?”) score on the SSI, would still be predictive of
leader emergence. In fact, previous research in
Feedback statements: the number of times similar types of groups indicates that there are
a participant provided feedback to another no significant differences in group productivity
member (e.g., “That’s good,” “I see”) or group satisfaction as a function of whether
the leader is male or female (Anderson & Blan-
Task-irrelevant statements: the number of chard, 1982; Anderson & Thacker, 1985; Gok-
statements made by each participant that tepe & Schneier, 1988).
were unrelated to the group problem-solv- To explore the impact of communication
ing task. skills/competence, extraversion, and talkative-
ness on leader emergence, we conducted a se-
These six coded variables allowed some mea- ries of discriminant analyses using the various
surement of both the quantity and the quality of indexes of verbal behaviors (amount of speak-
the verbal interaction in the group problem- ing, directive statements, etc.), EPI Extraver-
solving discussions. The first two variables sion scores, and total score on the SSI as pre-
were determined by simple word and speaking- dictors of group membership (leader or subor-
turn counts. Two trained coders independently dinate). As expected, there were strong positive
tallied the last four variables. Intercoder reli- correlations among the various indexes of ver-
abilities were high (all rs ⬎ .85). Differences in bal behaviors (e.g., amount of speaking, direc-
the two coders’ tallies of each variable were tive statements, feedback statements, and ques-
rectified by having the coders view together the tions asked correlated from .45 to .64). There-
discrepant videotape segments to agree on the fore, only total amount of speaking was used as
90 RIGGIO, RIGGIO, SALINAS, AND COLE

the indicator of verbal behavior. In addition, and Table 2

as expected, there was a strong positive corre- Classification of Leader Versus Subordinate
lation between extraversion and total score on Actual classification
the SSI (r ⫽ .47). Furthermore, there were Predicted
classification Leader Subordinate Total
significant positive correlations between both
extraversion and SSI total score and total Leader
n 15 24 39
amount of speaking (rs ⫽ .19 and .26, respec-
% 38.5 61.5
tively; both ps ⬍ .01). A stepwise discriminant Subordinate
analysis with the three predictor variables indi- n 20 159 179
cated that total amount of speaking entered the % 11.1 88.9
equation first, with extraversion second, and
total score on the SSI entering the equation last.
Both amount of speaking and extraversion con-
tributed significantly to the prediction of leader Discussion
emergence, with total SSI adding only a trivial
and nonsignificant amount to the equation (see The results of this initial study were not to-
Table 1). Combined, the three predictors did tally unexpected. First, research on gender dif-
only a fair job of correctly classifying leaders ferences in leader emergence suggests that men
(38%) but a good job of correctly classifying are more likely to emerge as leaders than
women, particularly in short-term groups and
subordinates (89%; see Table 2). An analysis of
situations where the tasks may be perceived as
whether speaking time was a mediator of the
masculine (Eagly & Karau, 1991). Such was
extraversion–leader emergence rating was non-
clearly the case in our study, in which the brief
significant, suggesting that extraversion and task concerned making decisions about desert
speaking time contributed independently to survival, which we assume to be a more stereo-
leader emergence. We assumed that the effects typically masculine task.
of extraversion were largely nonverbal, rather Consistent with the babble hypothesis, we
than verbal, in nature. found that sheer amount of talking was the best
A stepwise multiple regression analysis was predictor (besides gender) of who was chosen
conducted regressing the constructed “leader leader. Moreover, it did not appear that quality
potential score” on the three predictor variables. of communication had much effect, as evi-
The results almost perfectly paralleled the dis- denced by the negligible contribution of overall
criminant analysis. Only total amount of speak- communication skill/competence (total SSI
ing and extraversion significantly predicted the score) to leader emergence. We did find, how-
leader potential score, accounting for slightly ever, that extraversion had a significant, but
more than 15% of the variance, R2 ⫽ .39, smaller, and independent impact on leader
F ⫽ 49.08, p ⬍ .001, with total score on the SSI emergence, consistent with the results of a
adding only a trivial, nonsignificant prediction meta-analysis by Lord, DeVader, and Alliger
to the equation. The mediator analysis was also (1986).
nonsignificant. In short, we hypothesized that more socially
skilled and competent individuals would be
more likely to emerge as leaders in small, in-
teracting work groups than would persons lack-
Table 1 ing highly developed communication skills.
Stepwise Discriminant Analysis Predicting Leader However, this was not the case for several rea-
Emergence (Selection as Leader Vs. Subordinate)
sons: First, likely owing to the nature of our
Predictor variable F(1, 216) p task, men were much more often elected group
1. Total speaking 32.29 .0001 leaders than women. Second, the short duration
2. Extraversion (EPI) 11.12 .001 of the initial session before electing a leader
3. SSI total 2.63 .12 may have made it difficult, if not impossible, to
Note. Wilks’s lambda F(1, 216) ⫽ 19.9, p ⬍ .0001. EPI ⫽
adequately determine socially skilled from un-
Eysenck Personality Inventory; SSI ⫽ Social Skills skilled persons (although extraverted individu-
Inventory. als did tend to garner more leadership votes than
 LEADER COMMUNICATION SKILLS 91

did introverts). In sum, we expected group making suggestions about procedures, delegating
members to choose leaders who were “best responsibilities, and solving task-related prob-
qualified,” as operationalized by being commu- lems, and fewer relationship-focused behaviors.
nicatively competent. Instead, group members We hypothesized that in these longer, more
presumably chose leaders on the basis of “sur- involved, and more varied task situations,
face” characteristics (i.e., extraversion, who leader communication skills/competence would
spoke most frequently) and gender-role stereo- be an important variable in predicting leader
types about who might be more competent in effectiveness. In addition, in this study, we op-
performing the survival task (i.e., men). erationalized leader effectiveness in several
On the basis of these results, we conducted a ways. First, we looked at the group performance
second study specifically to examine the effects on the two tasks. This is a common operation-
of communication skill/competence on the ef- alization of leader effectiveness— group output.
fectiveness of leaders of small groups. In this Second, we had the group members rate their
study, rather than allowing leaders to emerge, leaders using the Leader Behavior Description
we preselected leaders according to their com- Questionnaire (Stogdill & Coons, 1957). Fi-
munication skills/competence, as measured by nally, we had trained coders view the video-
the SSI. taped group interactions and rate leader
performance.
Study 2
Method
In this second study, we wanted to experi-
mentally manipulate levels of communication Participants. Participants were 315 under-
skill/competence of group leaders, and we also graduate students (124 men and 191 women)
sought to broaden the situation by having the recruited from introductory psychology courses
leaders lead the groups in two different tasks. who participated in exchange for course credit.
Drawing on classic work on leadership behav- Measures. In initial large groups (25–50)
iors (e.g., Likert, 1967; Stogdill & Coons, students completed the original SSI and the EPI
1957), we sought to create leadership situations Extraversion and Neuroticism scales, and also
and tasks that would draw forth both relation- the Self-Monitoring Scale (SMS; Snyder, 1974,
ship-oriented and task-oriented leader behav- 1987) and the Rosenberg Self-Esteem Scale
iors. We used the crash-landing task from the (Rosenberg, 1965). The internal consistencies
first study because it represented a highly ver- of the SSI total score and EPI Extraversion and
bal, discussion-oriented, problem-solving situa- Neuroticism scales were good to acceptable (co-
tion in which group members’ input would be efficients ⫽ .94, .71, and .68, respectively).
important to help creatively solve the survival The SMS is an 18-item self-report measure
problem. Effective leader behaviors in this sit- with a true–false response scale. The SMS has
uation should involve such things as encourag- been extensively validated (see Snyder, 1987),
ing members to contribute to the discussion, and the internal consistency of our participants
communicating ideas effectively, regulating the on the SMS was acceptable (KR20 ⫽ .68).
flow of the discussion, keeping the group on The Rosenberg Self-Esteem Scale is a brief,
task, and moving the group toward a consensus. 10-item, self-report measure of general self-
We expected this task would elicit relationship- esteem. The measure has a Likert-type 4-point
oriented leader behaviors and, to a lesser extent, response scale with anchors of strongly agree
some task-oriented behaviors. and strongly disagree. It has been used exten-
The second task was a simulation of an as- sively in research, and the internal consistency
sembly line, in which group members used tools of the scale for our participants was very good
to assemble a number of intricate steel contrap- (␣ ⫽ .89).
tions. We expected that this task would require Procedures. Participants were told that this
leaders to be more directive and that there study was a validation study for the paper-and-
would be less discussion, with the leader taking pencil measures, and that their participation was
a more supervisory role. Effective leader per- complete and they had been given a small
formance in this task situation should elicit amount of course credit. To disassociate the
many task-oriented leader behaviors, such as measures from the second part of the experi-
92 RIGGIO, RIGGIO, SALINAS, AND COLE

ment, another experimenter was present who dium, and low socially skilled leaders all sig-
asked whether the students would like to par- nificantly different from one another on total
ticipate in a different, “videotaped study of SSI score. On the SSI deviation score, only the
group interaction.” This time, 7 participants low-skilled leaders differed significantly from
were signed up for each session. Because par- the other two groups. This suggests that in terms
ticipants were given a large amount of course of social skill balance, high- and medium-level
credit for participation in the second study, leaders did not differ significantly, except on
nearly all of the students signed up for the total SSI score. However, the low-skilled lead-
second session. Videotaped sessions began at ers were truly lower scoring, in terms of both
least 2 weeks following the initial session. total SSI score and having greater deviation
During the intervening 2 weeks, all partici- among the SSI dimensions.
pants’ SSIs were scored, and group leaders were Because some students did not show up for
selected for each group. Leaders were selected the videotaping session, some groups proceeded
by Ronald E. Riggio to represent three levels of with only 5 or, in rare instances, only 4 mem-
communication skill/competence (high, me- bers. No groups had fewer than 4 members or
dium, and low) on the basis of both total score more than 6 members. In cases in which all 7
on the SSI and the SSI deviation score. Rather members arrived, one member (a nonleader)
than use a firm cutoff score, the SSI scores for was taken aside and offered to participate in
all group members were examined, and a leader another, shorter (unrelated) experiment for the
was selected who was exceptionally high and same amount of course credit. Because of tech-
balanced, exceptionally low and imbalanced, or nical problems with the videotaping equipment,
a middle-scorer on the SSI. Attempts were some groups did not receive video ratings, and
made to choose an equal number of male and some groups failed to complete all of the mea-
female leaders and an equal representation of sures correctly, so there were some groups with
high-, medium-, and low-SSI scoring leaders. missing data. In all, there were 58 groups: 39
However, the relatively low number of male with 6 members (including the leader), 16
participants, coupled with the fact that there with 5 members, and 3 with 4 members.
were few men with high SSI scores, led to In this study, there were two separate group
disproportionately fewer high-SSI male leaders. tasks: the desert survival task used in the first
Alternate leaders were selected in case the pre- study and an assembly task. In the assembly
assigned leader did not show up. There were a task, group members were given a box of parts
total of 21 men and 37 women selected as group (different sizes of steel braces, nuts, and bolts)
leaders. A breakdown by gender of leader and and some tools (screwdrivers and wrenches),
SSI score is presented in Table 3, along with the presented with a model of an assembled com-
mean total SSI scores and mean SSI deviation ponent, and instructed to make as many dupli-
scores for the different groups. The leaders rep- cates of the assembled model as possible in 15
resented the desired distribution, with high, me- min. The order of the two tasks was alternated,

Table 3
Breakdown of Leaders by Gender and Social Skills Inventory (SSI) Category
SSI category
High Medium Low
Measure and
gender M SD n M SD n M SD n
SSI total score
Male 581.8 14.7 4 508.6 10.5 8 410.2 28.8 9
Female 597.3 26.8 18 528.8 19.8 11 443.3 21.7 8
Total 594.5 25.5 22 520.3 19.1 19 425.8 30.2 17
SSI deviation score
Male 56.8 13.8 4 53.4 11.3 8 79.3 19.0 9
Female 64.9 14.9 18 58.0 16.1 11 103.8 22.3 8
Total 63.5 14.7 22 56.1 14.2 19 90.8 23.6 17
 LEADER COMMUNICATION SKILLS 93

with half of the groups doing the assembly task table so that the video camera in the adjoining
first and half starting with the cognitive task. room could record all of the members at the
Survival task. As in the first study, partici- start of the session. The experimenter then pre-
pants who were starting with the survival task tended to randomly draw out the name of the
were greeted by the experimenter, who affixed group leader (but actually picked the predeter-
large nametags with participants’ first names to mined leader). The leader was instructed that he
their shirt and instructed them to take any seat. or she was “in charge” of the group and that he
Six chairs were arranged in a semicircle facing or she had responsibility for overseeing the as-
a video camera that was out of immediate sight sembly of a metal contraption, an assembled
in an adjoining room. The experimenter also model of which was on the table in front of
placed a slip of paper with each participant’s them. The metal contraption was made out of
first name into a box and pretended to draw out different-sized steel braces held together with
a name of the group’s leader. However, the nuts and bolts. It consisted of a rectangular base
experimenter simply read off the name of the made from four steel braces and two parallel
preselected leader. The leader was instructed that vertical bars, on which were attached two dia-
he or she was “in charge” of the group and was mond-shaped pieces. In the model, a stack of
responsible for leading the group and for record- pencils rested between the two diamond-shaped
ing and reporting on the group’s performance. pieces, and the model was referred to as a
Participants were given 10 min to work on “pencil holder.” On the floor beside the table
the survival task alone and were asked to inde- were two large boxes of metal braces and a
pendently rank-order the items in terms of their smaller box full of nuts and bolts. On the table
survival value. Participants were explicitly told were two crescent wrenches that fit the nuts and
that they were trying to get as close as possible two screwdrivers.
to the rankings given by survival experts. Af- The experimenter instructed the group that
ter 10 min, the experimenter instructed the their task was to make as many assembled cop-
leader to “take over” and lead the group to a ies of the pencil holders as possible from the
consensus, as in the first study. Groups were boxed components in 15 min. Groups were told
given 15 min to complete the task, and this that the experimenter would do a “quality in-
entire period was videotaped. At the end of the spection” of each of the assembled components,
discussion, the leader recorded the group’s to ensure that they were correctly assembled
ranking and was taken to an adjoining room for and that all of the bolts were tightened. Only
the videotaped presentation of the group’s rank- components that passed the quality inspection
ing as part of another, unrelated study. The would be counted toward the group’s output/
other participants waited for the leader’s return score. The experimenter then left the leader in
if the survival task was the first task completed, charge and left the room.
or, if the survival task was the second task, If the assembly task was the second task
group members were taken to separate testing completed, group members completed the
cubicles where they completed Version XII–R LBDQ ratings for their leader. Upon completion
of the Leader Behavior Description Question- of the two sessions, participants were told the
naire (LBDQ; Stogdill & Coons, 1957). The purposes of the experiment and were dismissed.
LBDQ is designed to assess the leader on the Leader effectiveness measures. There were
dimensions of initiating structure (task orienta- three separate measures of leader effectiveness.
tion) and showing consideration (relationship The first was the LBDQ ratings made by the
orientation). LBDQ scores on the Initiating leader’s group members. These were averaged
Structure and Consideration subscales were av- across the 3–5 group members and yielded a
eraged across all group members to obtain mean score for both initiating structure and consider-
leader ratings for each group. ation behaviors. High scores on this measure
Assembly task. Participants starting with indicate that group members observed the dis-
the assembly task were greeted by the experi- play of behaviors that are associated with leader
menter, who affixed a large nametag with each effectiveness (see Bass, 1990, chap. 24). Scores
participant’s first name to his or her shirt and led on the Initiating Structure subscale presumably
the group to a long table. The experimenter had represent more task-oriented aspects of leader
the members begin the task standing behind the effectiveness, whereas the Consideration sub-
94 RIGGIO, RIGGIO, SALINAS, AND COLE

scale represents leader effectiveness in relation- some practice sessions where raters shared their
ship orientation. Schriesheim and Kerr (1974) ratings and discussed discrepancies. Each rater
reported strong internal consistency for both the completed all of the ratings alone, and each
Initiating Structure (␣ ⫽ .78) and Consideration rater received a different, random order of vid-
(␣ ⫽ .90) subscales of the LBDQ. Because of eotapes for viewing, with two raters viewing the
an experimenter error, three groups did not survival tasks first and then the assembly tasks,
complete LBDQ ratings. and the other two raters receiving the reverse
A second measure of leader effectiveness ordering. Because of technical difficulties with
concerned the group’s performances on the two the videotaping procedures (primarily poor au-
tasks. For the survival task, group performance dio, but also some video camera problems),
was defined as the summed total of the absolute only 40 of the 58 groups were complete enough
values of the discrepancy on each item between to allow leader ratings.
the group’s final ranking of the survival value of Initially, there were 11 ratings for the sur-
each item and the survival experts’ ranking. vival task and 7 items for the assembly task.
Thus, a low score on this measure suggested Item analyses of the two scales, and subsequent
that the group had come close to “solving” the discussions with the raters, revealed that 2 items
survival problem correctly, and a high score for the survival task rating scale and 1 item from
indicated poor group performance on this prob- the assembly rating scale were problematic (i.e.,
lem-solving task. For the assembly task, the they had low item–total correlations; some con-
measure of group performance consisted of the fusion existed with the relation of the items to
number of components correctly assembled effective leader performance on the part of rat-
(i.e., passing the quality inspection). ers). These items were eliminated. The remain-
The third measure of leader effectiveness ing 9 items from the survival task rating scale
consisted of trained student raters evaluating the and 6 items from the assembly task scale had
videotaped performance of each group leader. A high internal consistencies (␣s ⫽ .95 and .92,
list of dimensions was generated to represent respectively), so the items were summed for
what were believed to be effective leadership each rating scale. Agreement among the four
behaviors for the group discussion/survival raters was high, with alpha coefficients of .87
task; a separate list was generated for the as- and .80 for the summed scores on the survival
sembly task. These items included (for the sur- task and assembly task leader ratings,
vival task) “Does the leader keep the group respectively.
focused on the topic of discussion?,” “Does the
leader make suggestions or contribute to the Results
discussion?,” “Does the leader establish rapport
with group members?,” “Does the leader en- Zero-order correlations among the leader ef-
courage discussion?,” and “Does the leader fectiveness ratings are presented in Table 4. As
make an effort to ensure group members’ agree- can be seen, the group members’ LBDQ scores
ment on any final decisions?” and (for the as- were, for the most part, positively and signifi-
sembly task) “Does the leader ‘take charge’ of cantly correlated with the videotape ratings
and supervise the assembly line?,” “Does the from the two group tasks. However, neither
leader communicate his or her ideas effectively measure of group performance—the discrep-
to the group?,” “Does the leader make sugges- ancy of the group’s rating from the expert rating
tions on how to best accomplish the task?,” and on the survival task nor the number of compo-
“Does the leader delegate responsibility?” All nents completed in the assembly task— corre-
items were rated using a 5-point scale with the lated significantly with any of the other leader-
anchors to little or no degree (1) to to a great ship effectiveness ratings.
degree (5). To examine the effect of communication/so-
A group of four students (two men and two cial skills on rated leader performance we con-
women) was trained to rate each leader’s effec- ducted a series of multivariate analyses of vari-
tiveness using the rating scales. The training ance (MANOVAs) systematically combining
consisted of a group discussion of what consti- the measures of leader effectiveness, and then
tuted effective leadership behaviors, a detailed followed up with univariate analyses of vari-
explanation of the two experimental tasks, and ance (ANOVAs) and post hoc contrasts.
 LEADER COMMUNICATION SKILLS 95

Table 4
Zero-Order Correlations Among Leader Performance/Effectiveness Measures
Measure 1 2 3 4 5 6
1. LBDQ–IS — .69** .56** .42** ⫺.06 .07
(55) (40) (40) (55) (55)
2. LBDQ–C — .47** .25 ⫺.03 .22
(40) (40) (55) (55)
3. Discussion rating — .58** ⫺.20 .11
(40) (40) (40)
4. Assembly rating — ⫺.13 .20
(40) (40)
5. Survival performance — .06
(58)
6. Assembly performance —
Note. Values in parentheses are the n for that coefficient. LBDQ ⫽ Leadership Behavior
Description Questionnaire; IS ⫽ Initiating Structure subscale; C ⫽ Consideration subscale.
** p ⬍ .01.

The first analysis looked for gender differ- effectiveness measures were either significant
ences in leader performance. A MANOVA us- or approached significance on the univariate
ing all six measures of leader effectiveness tests. It is important to note that the number of
(LBDQ Initiating Structure subscale, LBDQ groups with complete data on all six measures
Consideration subscale, video ratings from the of leader effectiveness was low because only 40
two tasks, number of items assembled, and dis- groups received video ratings. Therefore, we
crepancy from expert ranking of survival task) conducted separate univariate ANOVAs for the
indicated no significant gender differences LBDQ ratings and video ratings to include as
across the six measures: Wilks’s lambda, F(6, many groups as possible in the analyses. These
33) ⫽ 1.15, ns. Moreover, none of the univari- results showed significant effects for the LBDQ
ate Fs were significant, although leader ratings Initiating Structure and Consideration sub-
on the assembly task approached significance, scales, Fs(2, 52) ⫽ 4.25 and 4.70, respectively,
with women leaders receiving slightly, but non- ps ⬍ .02, and for the video rating of the survival
significantly, higher ratings than male leaders, task discussion, F(2, 37) ⫽ 5.04, p ⬍ .02, but
F(1, 38) ⫽ 2.56, p ⬍ .12. In addition, a 2 not for the video rating of the assembly task
(gender of leader) ⫻ 3 (level of leader commu- (F ⬍ 1.0). Post hoc comparisons suggested that
nication/social skill) MANOVA was run to ex- there were marked differences between the
amine for possible interaction effects, and there high- and low-SSI leaders, with marginally sig-
was no significant interaction between leader nificant differences between the medium- and
social/communication skills and gender, Wilks’s low-SSI leaders for these three leader effective-
lambda, F(4, 64) ⫽ 1.54, p ⫽ .20, in the overall ness ratings. Means and standard deviations for
MANOVA or in the univariate tests. Because all groups on the six measures of leader effec-
there were no significant gender differences, tiveness/performance are presented in Table 5,
gender was dropped from further analyses. along with significant post hoc comparisons.
A second MANOVA looked at differences Finally, in order to explore both how the SSI
among leaders with high, medium, and low subscale dimensions and the other measures of
social communication skills on all six measures personality (extraversion, self-monitoring, and
of leader effectiveness. The multivariate test self-esteem) play a part in leader effectiveness,
approached significance: Wilks’s lambda, F(12, Table 6 presents the correlations between these
64) ⫽ 1.67, p ⬍ .10. Univariate tests showed various individual-difference measures and the
that there were no significant differences on six measures of leader effectiveness. However,
either performance measure (number of items it is important to interpret these relationships
assembled or discrepancy of group’s from ex- cautiously because leaders were not randomly
pert’s ranking on the survival task; both assigned. As can be seen, several SSI sub-
Fs ⬍ 1.0). However, the remaining four leader scales—particularly those related to emotional
96 RIGGIO, RIGGIO, SALINAS, AND COLE

Table 5
Means and Standard Deviations of Measures of Leader
Effectivness/Performance by Social Skills Inventory (SSI) Category
SSI category
High Medium Low
Measure M SD M SD M SD
LBDQ–IS 59.0a,b 7.9 52.4b 8.4 50.7a 11.3
LBDQ–C 78.0a 5.5 73.9 5.9 71.2a 8.9
Survival video rating 130.3a,b 20.1 107.1b 27.6 99.8a 32.0
Assembly video rating 56.3 19.0 56.7 20.9 48.8 16.3
Survival task performance 69.6 12.4 69.0 11.5 70.4 14.0
Assembly task performance 6.4 0.5 6.6 0.5 6.5 0.5
Note. Within each row, subscripts indicate that means are significantly different from each
other at p ⬍ .05 (subscript a) or p ⬍ .10 (subscript b). LBDQ ⫽ Leadership Behavior
Description Questionnaire; IS ⫽ Initiating Structure subscale; C ⫽ Consideration subscale.

and social expressiveness and social control (a play a part in leader effectiveness/performance
form of social role-playing ability)—along with in small, interacting task groups. Specifically,
extraversion, self-monitoring, and, to a lesser leaders who were better and more skilled com-
extent, self-esteem, were significantly posi- municators were rated as more effective leaders,
tively correlated with both group member and by both the leaders’ group members and third-
video observer ratings of leader effectiveness. party observers, than were nonskilled commu-
These relationships were stronger, however, for nicators. However, this difference was more
the survival discussion task than for the assem- pronounced in the discussion-based desert sur-
bly task. Not surprisingly, there were few sig- vival task than it was in the hands-on assembly
nificant correlations between individual-differ- task. Contrary to expectations, socially skilled
ence measures and the two measures of group leaders did not have better performing groups as
performance. assessed by our measures of group perfor-
mance. It is also important to note that the group
Discussion performance measures did not correlate signif-
icantly with any of the leadership effectiveness
There was partial support for our main hy- ratings (or with each other). In any case, we did
pothesis that communication/social skills would not find evidence that leader communication/

Table 6
Zero-Order Correlations Between Individual-Difference Measures and Leader Effectiveness
Measure LBDQ–IS LBDQ–C Discussion rating Assembly rating Survival Assembly performance
SSI–EE .32** .42* .23 .18 .01 .20
SSI–ES .19 .19 .34* .14 ⫺.06 .16
SSI–EC .08 .24 .40** ⫺.01 ⫺.06 .15
SSI–SE .32* .33* .42** .28 .00 .24
SSI–SS ⫺.07 .14 .02 ⫺.19 .06 ⫺.04
SSI–SC .46** .42** .46** .24 .04 .15
SSI total .34* .42** .46** .18 .00 .22
SSI deviation ⫺.36** ⫺.34* ⫺.28 ⫺.13 ⫺.03 ⫺.19
Extraversion .32* .33* .44** .24 ⫺.03 .26*
Neuroticism ⫺.15 ⫺.06 ⫺.03 .09 .12 ⫺.01
Self-monitoring .20 .40** .35* .10 .01 .15
Self-esteem .23 .22 .36* .32* ⫺.28* .20
Note. LBDQ ⫽ Leadership Behavior Description Questionnaire; IS ⫽ Initiating Structure subscale; C ⫽ Consideration
subscale; SSI ⫽ Social Skills Inventory; EE ⫽ Emotional Expressivity; ES ⫽ Emotional Sensitivity; EC ⫽ Emotional
Control; SE ⫽ Social Expressivity; SS ⫽ Social Sensitivity; SC ⫽ Social Control.
* p ⬍ .05. ** p ⬍ .01.
 LEADER COMMUNICATION SKILLS 97

social skills significantly impacted either of our components of emotional and social intelli-
measures of group task performance. gence); related personality constructs of extra-
The finding that leader communication/social version, self-monitoring, and self-esteem/confi-
skills have a greater impact on rated evaluations dence; and leader performance, although these
of leader effectiveness than on group perfor- correlations must be interpreted cautiously be-
mance was not wholly unexpected. For exam- cause of our nonrandom leader selection. As
ple, a review of research on personality and expected, the strongest relationships were be-
leader effectiveness suggests that there is more tween the expressiveness subscales of the SSI
evidence that certain leader personality vari- (Emotional Expressivity and Social Expressiv-
ables (e.g., sociable, active, emotionally bal- ity) and social role-playing skill (Social Con-
anced, conscientious) have a positive effect on trol). With only one exception, these three SSI
rated leader effectiveness than evidence of per- subscales were significantly positively corre-
sonality effects on team performance (Hogan, lated with group members’ LBDQ ratings and
Curphy, & Hogan, 1994). Research on charis- with observer ratings of leader performance in
matic and transformational leadership also sug- the survival task discussion. This makes sense
gests, for example, that transformational leaders because of the nature of the discussion task,
have stronger effects on subordinate ratings of which, in contrast to the assembly task, gave
leader effectiveness than on measures of group leaders greater opportunity to demonstrate their
performance (Lowe et al., 1996). communication skills. Moreover, in brief group
Moreover, it was not surprising that leader interactions, a leader’s expressive and role-
communication skills had greater impact in the playing skills are those dimensions of commu-
discussion-based survival task than in the as- nication skill that will be most visible. It would
sembly task. Informal observation of group per- be much more difficult to determine whether a
formance showed that the leader spent a greater leader were emotionally and socially sensitive
proportion of time verbally communicating and and whether that leader had good control over
regulating the flow of communication in the his or her emotions (Emotional Control) in a
discussion task than in the assembly task. In brief encounter, although it is interesting to note
fact, the assembly task appeared more chaotic that there were significant positive correlations
and had less overall verbal communication between both the Emotional Sensitivity and
(long periods would pass with virtually no com- Emotional Control SSI subscales and observer
munication because all members were busily ratings of leaders in the survival discussion task.
engaged in assembly), and raters later noted that In addition, consistent with previous research,
the leaders played less of an obvious “leader- leaders who had high imbalances in their com-
ship role” in the assembly task. munication/social skills, as indicated by the SSI
Interestingly, we did not find any significant deviation score, received significantly lower
gender differences in leader performance, sug- leader effectiveness ratings from group mem-
gesting that male and female leaders performed bers. Finally, personality constructs that are
similarly, despite the fact that there was such a both theoretically and empirically related to so-
large gender difference in leader emergence in cial skills and social intelligence—namely self-
the first study, and despite the fact that both the monitoring, extraversion, and self-esteem—also
survival task and the assembly task are stereo- tended to be positively correlated with rated
typically masculine tasks. If there had been gen- leader effectiveness, although the magnitude
der differences in leader effectiveness, we and consistency of these correlations tended to
might have expected female leaders to perform be less than for the SSI subscales.
better because research on the SSI, as well as One strength of this study was the multiple
research on emotional communication skills, perspectives on leader effectiveness/perfor-
suggests that women have both higher overall mance. We included group member ratings on a
scores on the SSI and greater levels of emo- tried and true measure, the LBDQ, observer
tional encoding and decoding skill than do men ratings on a measure of our own design (but one
(see Hall, 1984; Riggio & Carney, 2003). that appeared to have good psychometric prop-
We also explored the relationship between erties), and measures of actual group perfor-
the various SSI dimensions, representing both mance. Unfortunately, our weakest measures
emotional and social communication skills (i.e., seemed to be the group performance measures.
98 RIGGIO, RIGGIO, SALINAS, AND COLE

There were also some limitations to this Method

study. First, there were some imbalances in both
the gender of participants in the experiment and Participants. Participants were 76 members
the distribution of selected leaders, with many of the fire departments from two cities and one
more women participating and very few men county in Southern California who held the
scoring high enough on the SSI to be included supervisory positions of captain (n ⫽ 62), bat-
as high socially skilled leaders. Part of this was talion chief (n ⫽ 11), or division chief (n ⫽ 3).
due to using an introductory psychology partic- These supervisory positions represent the up-
ipant pool that was more than two thirds ward chain of command in the fire service. In
women. We were also hampered by technical addition, 173 firefighters holding “line” posi-
difficulties so that we could not obtain observer tions of firefighter or engineer participated by
ratings on all of our groups, although the high evaluating the performance of their supervisors
or captains.
intercorrelations among our leadership effec-
Procedure. All participants completed both
tiveness ratings and the consistency of the sig-
the full version of the SSI and the Satisfaction
nificant findings both suggest that this was not a
With Supervisor subscale of the Job Descriptive
major problem. As always, one should also be Index (JDI; Smith, Kendall, & Hulin, 1969) at
cautious in generalizing results obtained from their places of employment. In separate ses-
temporary groups of students performing in a sions, participants who were line employees
laboratory-based simulation, but it is important (e.g., firefighters) rated their immediate super-
to note that our participants seemed to take the visors (captains) using a 10-item supervisory
tasks, and their leadership roles, very seriously. performance rating scale that was adapted from
the actual performance appraisal instrument
Study 3 used by the county fire department. Items on
this performance rating instrument used a
Obvious limitations of the first two studies 5-point rating scale with anchors of 1 (unsatis-
were the contrived nature of the laboratory set- factory) to 5 (outstanding). Specific items as-
ting and the use of student leaders and follow- sessed were work productivity, work planning,
ers. In a third study, we sought to generalize work organization, delegation, employee devel-
these results to an actual leadership setting, ex- opment, leadership ability, supervisory control,
ploring primarily the role that social and emo- general decision making, fire-ground decision
tional communication skills play in rated lead- making, and medical emergency decision mak-
ing. Intercorrelations among the 10 items were
ership performance of work-group supervisors.
high for all sets of ratings (rs ⫽ .37–.77; scale
For this study, we examined supervisors and
␣ ⫽ .92), indicating good internal consistency.
leaders in the fire service. We focused on two
Therefore, the 10 items were summed into an
supervisory levels: captains in the fire service, overall performance rating for each supervisor
who are the first supervisory line, and individ- rated.
uals who have higher leadership roles, battalion Using the same performance rating instru-
and division chiefs. Although we could not spe- ment, captains rated their battalion chiefs’ per-
cifically address leader emergence in this pop- formance. Battalion chiefs rated both their sub-
ulation, we hypothesized that social and emo- ordinates’ (captains’) performance and the per-
tional communication skills would be important formance of their superiors (the division chiefs).
to supervisory performance. Moreover, we ex- The result was both (a) superior and subordinate
pected that as one moved up the chain of com- ratings of performance for captains and (b) sub-
mand, the importance of communication skills ordinate ratings of performance for all chiefs.
would increase, because even early research on Within-rater correlations were quite high for all
managerial communication suggests that al- performance ratings (range ⫽ .36 to .81; median
though front-line supervisors spend about half r ⫽ .58). Subordinates’ ratings of supervisor
of their time engaged in communication, top- performance were thus aggregated and aver-
level executives and leaders spend nearly all of aged for each supervisor. Within-rater correla-
their time involved in communication activities tions using the JDI Satisfaction With Supervisor
(Klemmer & Snyder, 1972; Stewart, 1967). subscale were lower but still acceptable
 LEADER COMMUNICATION SKILLS 99

(range ⫽ .18 to .67; median r ⫽ .39). Therefore, groups of college students. When allowed to
these JDI ratings were also aggregated to create select their own group leaders, participants
a mean satisfaction score for each supervisor. chose men over women, selected extraverts
Because of the small number of division chiefs, over introverts, and tended to pick the person
these were combined with the battalion chiefs who talked the most, seemingly without regard
and analyzed together. to the “quality” of their verbal participation in a
discussion-based, problem-solving task. These
Results and Discussion results are consistent with the babble hypothesis
(Bass, 1990) and with research in social percep-
Contrary to expectations, captains’ scores on tion that suggests that highly “visible” individ-
the SSI were not correlated with either subor- uals receive more attention and may be more
dinate ratings of captains’ performance (r ⫽ positively evaluated in brief, initial encounters
⫺.01, ns) or superior ratings of captains’ per- (Schneider, Hastorf, & Ellsworth, 1979). De-
formance (r ⫽ ⫺.14, ns). However, there was a spite these somewhat disappointing results, we
strong significant correlation between subordi- would expect that in longer interactions, where
nate ratings of supervisory performance and the group members have greater opportunity to
JDI Satisfaction With Supervision subscale evaluate one another’s behavior and contribu-
(r ⫽ .75, p ⬍ .001). For chiefs, however, total tions to the interaction, communication/social
SSI score was significantly positively correlated skills would play a much greater part in leader
with both subordinates’ ratings of performance emergence.
for chiefs (r ⫽ .61, p ⬍ .01) and scores on the The results obtained in our second study, of
JDI Satisfaction With Supervision subscale leader effectiveness, were more supportive of
(r ⫽ .89, p ⬍ .001). our hypotheses. Leaders who were selected by
Because of the differences in patterns of cor- virtue of their possession of well-developed
relations between the SSI and measures of per- skills in both emotional and social communica-
formance and satisfaction for captains and tion received more positive evaluations of their
chiefs, it seemed important to look at mean leadership effectiveness from both their group
differences on the SSI for the three levels of fire members and third-party observers. However,
service employees: chiefs, captains, and fire- they did not necessarily lead better performing
fighters. The mean total SSI scores for the three groups. Moreover, as might be expected, lead-
groups, respectively, were 500.8, 490.6, and ers’ communication/social skills seemed more
487.2 (all standard deviations were about 50.0). important in the group discussion task than in
Although a one-way ANOVA indicated that the hands-on assembly task. Clearly, the lead-
there were no statistically significant differ- ership situation plays a very important part in
ences among the groups, the means were in the dictating the importance of a leader’s commu-
predicted direction, with social skills scores ris- nication skills in successful leadership (see, e.g.,
ing as one moves up the chain of command. Fiedler, 1993). However, it seems obvious that
These results provide modest support for the a leader’s communication skills, as broadly de-
general assertion that supervisor communica- fined by the model represented by the SSI (Rig-
tion/social skills play a role in determining suc- gio, 1986, 1989), including such elements as
cessful supervisor performance and subordinate social role-playing skill and the ability to regu-
satisfaction with that supervision. However, this late and control emotional displays, would be an
was true only for upper-level supervisors effective component of leadership across a wide
(chiefs), not for line supervisors (captains). variety of leadership situations.
Zaccaro (2002) suggested that there has been
General Discussion a resurgence in the trait approach to studying
leadership— but with a new twist. The leader-
The first two studies were conducted to ex- ship traits that are being investigated are com-
plore the importance of emotional and social plex. They involve characteristics such as be-
communication skills on leader emergence and havioral flexibility, creativity, and cognitive
effectiveness in small groups. Contrary to our complexity— characteristics that represent
expectations, communication skill played little complex skills or abilities rather than more tra-
part in leader emergence in our interacting ditional personality traits. Among these impor-
100 RIGGIO, RIGGIO, SALINAS, AND COLE

tant leader characteristics are social and emo- and job performance. Basic and Applied Social
tional skills and abilities, such as the communi- Psychology, 6, 345–361.
cation/social skills that were examined in the Bass, B. M. (1949). An analysis of the leaderless
present studies. These are being identified as group discussion. Journal of Applied Psychol-
ogy, 33, 527–533.
components of both social and emotional intel-
Bass, B. M. (1990). Bass & Stogdill’s handbook of
ligence (see Riggio, Murphy, & Pirozzolo, leadership (3rd ed.). New York: Free Press.
2002). The studies presented here represent a Bass, B. M. (2002). Cognitive, social, and emotional
first step in examining the relationship of basic intelligence of transformational leaders. In R. E.
communication/social skills to both leader Riggio, S. E. Murphy, & F. J. Pirozzolo (Eds.),
emergence and leader effectiveness. Additional Multiple intelligences and leadership (pp. 105–
studies exploring these relationships in different 118 ). Mahwah, NJ: Erlbaum.
leadership situations, such as leaders working Bray, D. W., Campbell, R. J., & Grant, D. L. (1974).
under stressful conditions, have been somewhat Formative years in business: A long-term AT&T
consistent with the results obtained here. For study of managerial lives. New York: Wiley.
example, in one study it was found that socially Caruso, D. R., Mayer, J. D., & Salovey, P. (2002).
Emotional intelligence and emotional leadership.
skilled leaders performed better under both In R. E. Riggio, S. E. Murphy, & F. J. Pirozzolo
stressful and nonstressful conditions (Halver- (Eds.), Multiple intelligences and leadership (pp.
son, Murphy, & Riggio, 2002). 55–74 ). Mahwah, NJ: Erlbaum.
Leadership is a complex construct. The char- Chemers, M. M. (1997). An integrative theory of
acteristics that contribute to effective leadership leadership. Mahwah, NJ: Erlbaum.
are complex and multifaceted. Basic skills in Dansereau, F., Graen, G., & Haga, B. (1975). A
emotional and nonverbal communication and vertical dyad linkage approach to leadership within
skills relating to effective communication in formal organizations: A longitudinal investigation
social situations (social role-playing skills, so- of the role making process. Organizational Behav-
cial sensitivity, etc.) are likely some of these ior and Human Performance, 13, 46 –78.
complex and critical characteristics of effective DePaulo, B. M., & Friedman, H. S. (1998). Nonver-
bal communication. In D. T. Gilbert, S. T. Fiske, &
leaders. By assessing possession of these basic G. Lindzey (Eds.), The handbook of social psy-
social abilities, through self-report methods as chology (Vol. 2, pp. 3– 40). New York: McGraw-
done here or, more important, using perfor- Hill.
mance-based assessments of these skills (see DePaulo, P. J. (1992). Applications of nonverbal
Riggio, 1992; Riggio & Riggio, 2001), we can behavior research in marketing and management.
better understand some of the critical character- In R. S. Feldman (Ed.), Applications of nonverbal
istics that are important for leaders. This will, of behavioral theories and research (pp. 63– 87).
course, have important implications for the de- Hillsdale, NJ: Erlbaum.
velopment of leadership theory—particularly Dobbins, G. H., Long, W. S., Dedrick, E. J., &
theories that emphasize leadership as a social Clemons, T. C. (1990). The role of self-monitoring
and gender on leader emergence: A laboratory and
relationship. Moreover, we will gain insight
field study. Journal of Management, 16, 609 – 618.
into how to more efficiently and successfully Eagly, A. H., & Karau, S. J. (1991). Gender and the
train and develop current and future leaders. emergence of leaders: A meta-analysis. Journal of
Personality and Social Psychology, 60, 685–710.
Ellis, R. J. (1988). Self-monitoring and leadership
References emergence in groups. Personality and Social Psy-
chology Bulletin, 14, 681– 693.
Aditya, R. N., & House, R. J. (2002). Interpersonal Ellis, R. J., Adamson, R. S., Deszca, G., & Cawsey,
acumen and leadership across cultures: Pointers T. F. (1988). Self-monitoring and leadership emer-
from the GLOBE study. In R. E. Riggio, S. E. gence. Small Group Behavior, 19, 312–324.
Murphy, & F. J. Pirozzolo (Eds.), Multiple intelli- Eysenck, H., & Eysenck, S. (1968). Manual for the
gences and leadership (pp. 215–240). Mahwah, Eysenck Personality Inventory. San Diego, CA:
NJ: Erlbaum. Educational and Industrial Testing Service.
Anderson, L. R., & Blanchard, P. N. (1982). Sex Fiedler, F. E. (1993). The leadership situation and the
differences in task and social-emotional behavior. black box in contingency theories. In M. M. Chem-
Basic and Applied Social Psychology, 3, 109 –139. ers & R. Ayman (Eds.), Leadership theory and
Anderson, L. R., & Thacker, J. (1985). Self-monitor- research: Perspectives and directions (pp. 1–28).
ing and sex as related to assessment center ratings San Diego, CA: Academic Press.
 LEADER COMMUNICATION SKILLS 101

Finkle, R. B. (1976). Managerial assessment centers. Howard, A., & Bray, D. W. (1990). Predictions of
In M. D. Dunnette (Ed.), Handbook of industrial managerial success over long periods of time: Les-
and organizational psychology (pp. 861– 888). sons from the management progress study. In K. E.
Chicago: Rand McNally. Clark & M. B. Clark (Eds.), Measures of leader-
Fletcher, C. (1989). Impression management in the ship (pp. 113–130). West Orange, NJ: Leadership
selection interview. In R. A. Giacalone & P. Library of America.
Rosenfeld (Eds.), Impression management in the Howell, J. M., & Frost, P. J. (1989). A laboratory
organization (pp. 269 –281). Hillsdale, NJ: Erlbaum. study of charismatic leadership. Organizational
Friedman, H. S., Prince, L. M., Riggio, R. E., & Behavior and Human Decision Processes, 43,
DiMatteo, M. R. (1980). Understanding and as- 243–269.
sessing nonverbal expressiveness: The Affective Kahn, R., & Katz, D. (1952). Leadership practices in
Communication Test. Journal of Personality and relation to productivity and morale. Ann Arbor:
Social Psychology, 39, 333–351. University of Michigan, Institute for Social Research.
Friedman, H. S., Riggio, R. E., & Casella, D. (1988). Kanter, R. M. (1983). The change masters. New
Nonverbal skill, personal charisma, and initial at- York: Simon & Schuster.
traction. Personality and Social Psychology Bulle- Katz, D., & Kahn, R. (1966). The social psychology
tin, 14, 203–211. of organizations. New York: Wiley.
Gardner, W. L., & Avolio, B. J. (1998). The charis- Kenny, D. A., & Zaccaro, S. J. (1983). An estimate of
matic relationship: A dramaturgical perspective. variance due to traits in leadership. Journal of
Academy of Management Review, 23, 32–58. Applied Psychology, 68, 678 – 685.
Gifford, R., Ng, C. F., & Wilkinson, M. (1985). Klemmer, E. T., & Snyder, F. W. (1972). Measure-
Nonverbal cues in the employment interview: ment of time spent communicating. Journal of
Links between applicant qualities and interviewer Communication, 22, 142–158.
judgments. Journal of Applied Psychology, 70, Leary, M. R. (1989). Self-presentational processes in
729 –736. leadership emergence and effectiveness. In R. A.
Goktepe, J. R., & Schneier, C. E. (1988). Sex and
Giacalone & P. Rosenfeld (Eds.), Impression man-
gender effects in evaluating emergent leaders in
agement in the organization (pp. 363–374). Hills-
small groups. Sex Roles, 19, 29 –36.
dale, NJ: Erlbaum.
Goleman, D. (1998a, November–December). What
Likert, R. (1967). The human organization. New
makes a good leader? Harvard Business Re-
York: McGraw-Hill.
view, 76, 93–102.
Lord, R. G., DeVader, C. L., & Alliger, G. M. (1986).
Goleman, D. (1998b). Working with emotional intel-
ligence. New York: Bantam. A meta-analysis of the relation between personal-
Graen, G., Novak, M., & Sommerkamp, P. (1982). ity traits and leadership perceptions: An applica-
The effects of leader-member exchange and job tion of validity generalization procedures. Journal
design on productivity and satisfaction: Testing a of Applied Psychology, 71, 402– 410.
dual attachment mode. Organizational Behavior Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N.
and Human Performance, 30, 109 –131. (1996). Effectiveness correlates of transforma-
Hall, J. A. (1984). Nonverbal sex differences: Com- tional and transactional leadership: A meta-ana-
munication accuracy and expressive style. Balti- lytic review of the MLQ literature. Leadership
more: Johns Hopkins University Press. Quarterly, 7, 385– 425.
Halverson, S., Murphy, S. E., & Riggio, R. E. (2002). Mann, R. D. (1959). A review of the relationships
Charismatic leadership under crisis situation: The between personality and performance in small
role of social and emotional intelligence. Manu- groups. Psychological Bulletin, 56, 241–270.
script submitted for publication. Mayer, J. D., & Salovey, P. (1997). In P. Salovey &
Hogan, R., Curphy, G. J., & Hogan, J. (1994). What D. Sluyter (Eds.), Emotional development and
we know about leadership: Effectiveness and per- emotional intelligence: Implications for educators
sonality. American Psychologist, 49, 493–504. (pp. 3–31). New York: Basic Books.
Hollander, E. P. (1985). Leadership and power. In G. Mintzberg, H. (1973). The nature of managerial
Lindzey & E. Aronson (Eds.), The handbook of work. New York: Harper & Row.
social psychology (3rd ed., pp. 485–537). New Mullen, B., Salas, E., & Driskell, J. E. (1989). Sa-
York: Random House. lience, motivation, and artifact as contributions to
House, R. J. (1971). A path-goal theory of leader the relation between participation rate and leader-
effectiveness. Administrative Science Quarterly, 1, ship. Journal of Experimental Social Psychol-
321–338. ogy, 25, 545–559.
House, R. J., & Mitchell, T. (1974). Path-goal theory Mumford, M. D., Zaccaro, S. J., Johnson, J. F.,
of leadership. Journal of Contemporary Busi- Diana, M., Gilbert, J. A., & Threlfall, K. V. (2000).
ness, 3, 81–98. Patterns of leader characteristics: Implications for
102 RIGGIO, RIGGIO, SALINAS, AND COLE

performance and development. Leadership Quar- in hiring interviews. Journal of Applied Social
terly, 11, 115–133. Psychology, 18, 331–348.
Murphy, S. M. (2002). Leader self-regulation: The Riggio, R. E., Throckmorton, B., & DePaola, S.
role of self-efficacy and multiple intelligences. In (1990). Social skills and self-esteem. Personality
R. E. Riggio, S. E. Murphy, & F. J. Pirozzolo and Individual Differences, 11, 799 – 804.
(Eds.), Multiple intelligences and leadership (pp. Riggio, R. E., Tucker, J., & Coffaro, D. (1989).
163–186). Mahwah, NJ: Erlbaum. Social skills and empathy. Personality and Indi-
Perez, J. E., Riggio, R. E., & Kopelowicz, A. (2002). vidual Differences, 10, 93–99.
Social skill imbalances as indicators of psycho- Riggio, R. E., Watring, K., & Throckmorton, B.
pathology: An exploratory investigation. Manu- (1993). Social skills, social support, and psycho-
script submitted for publication. social adjustment. Personality and Individual Dif-
Riggio, R. E. (1986). Assessment of basic social ferences, 15, 275–280.
skills. Journal of Personality and Social Psychol- Riggio, R. E., & Zimmerman, J. A. (1991). Social
ogy, 51, 649 – 660. skills and interpersonal relationships: Influences
Riggio, R. E. (1987). The charisma quotient. New on social support and support seeking. In W. H.
York: Dodd, Mead. Jones & D. Perlman (Eds.), Advances in personal
Riggio, R. E. (1989). Manual for the Social Skills relationships (Vol. 2, pp. 133–155). London: Jes-
Inventory. Palo Alto, CA: Consulting Psycholo- sica Kingsley Press.
gists Press. Rosenberg, M. (1965). Society and the adolescent
Riggio, R. E. (1992). Social interaction skills and self image. Princeton, NJ: Princeton University
nonverbal behaviors. In R. S. Feldman (Ed.), Ap- Press.
plications of nonverbal behavioral theories and Rosenthal, R., Hall, J. A., DiMatteo, M. R., Rogers,
research (pp. 3–30). Hillsdale, NJ: Erlbaum. P. L., & Archer, D. (1979). Sensitivity to nonverbal
Riggio, R. E. (1998). Charisma. In H. S. Friedman communication: The PONS test. Baltimore: Johns
(Ed.), Encyclopedia of mental health (pp. 387– Hopkins University Press.
396). San Diego, CA: Academic Press.
Schneider, D. J., Hastorf, A. H., & Ellsworth, P. C.
Riggio, R. E. (2001). Interpersonal sensitivity re-
(1979). Person perception. Reading, MA: Addi-
search and organizational psychology: Theoretical
son-Wesley.
and methodological applications. In J. A. Hall & F.
Schriesheim, C. A., & Kerr, S. (1974). Psychometric
Bernieri (Eds.), Measurement of interpersonal sen-
properties of the Ohio State leadership scales. Psy-
sitivity. Mahwah, NJ: Erlbaum.
chological Bulletin, 81, 756 –765.
Riggio, R. E. (2002). Multiple intelligences and lead-
Smith, P. C., Kendall, L. M., & Hulin, C. L. (1969).
ership: An overview. In R. E. Riggio, S. E. Mur-
phy, & F. J. Pirozzolo (Eds.), Multiple intelli- The measurement of satisfaction in work and re-
gences and leadership (pp. 1– 6). Mahwah, NJ: tirement. Chicago: Rand-McNally.
Erlbaum. Snyder, M. (1974). The self-monitoring of expressive
Riggio, R. E., & Carney, D. (2003). Manual for the behavior. Journal of Personality and Social Psy-
Social Skills Inventory (2nd ed.). Redwood City, chology, 30, 526 – 637.
CA: MindGarden. Snyder, M. (1987). Public appearances/private real-
Riggio, R. E., & Friedman, H. S. (1982). The inter- ities: The psychology of self-monitoring. New
relationships of self-monitoring factors, personal- York: W. H. Freeman.
ity traits, and nonverbal social skills. Journal of Sorrentino, R. M., & Boutillier, R. G. (1975). The
Nonverbal Behavior, 7, 33– 45. effect of quantity and quality of verbal interaction
Riggio, R. E., Messamer, J., & Throckmorton, B. on ratings of leadership ability. Journal of Exper-
(1991). Social and academic intelligence: Concep- imental Social Psychology, 11, 403– 411.
tually distinct but overlapping constructs. Person- Spitzberg, B. H., & Cupach, W. R. (1989). Handbook
ality and Individual Differences, 12, 695–702. of interpersonal competence research. New York:
Riggio, R. E., Murphy, S. E., & Pirozzolo, F. J. Springer-Verlag.
(Eds.). (2002). Multiple intelligences and leader- Stewart, R. (1967). Managers and their jobs: A study
ship. Mahwah, NJ: Erlbaum. of the similarities and differences in the way man-
Riggio, R. E., & Riggio, H. R. (2001). Self-report agers spend their time. London: Macmillan.
measurement of interpersonal sensitivity. In J. A. Stogdill, R. M. (1974). Handbook of leadership. New
Hall & F. Bernieri (Eds.), Measurement of inter- York: Free Press.
personal sensitivity. Mahwah, NJ: Erlbaum. Stogdill, R. M., & Coons, A. E. (Eds.). (1957).
Riggio, R. E., & Throckmorton, B. (1988). The rel- Leader behavior: Its description and measure-
ative effects of verbal and nonverbal behavior, ment. Columbus: Ohio State University, Bureau of
appearance, and social skills on evaluations made Business Research.
 LEADER COMMUNICATION SKILLS 103

Thornton, G. C., III (1992). Assessment centers in Zaccaro, S. J., Foti, R. J., & Kenny, D. A. (1991).
human resource management. Reading, MA: Ad- Self-monitoring and trait-based variance in leader-
dison-Wesley. ship: An investigation of leader flexibility across
Tucker, J. S., & Riggio, R. E. (1988). The role of multiple group situations. Journal of Applied Psy-
social skills in encoding posed and spontaneous chology, 76, 308 –315.
facial expression. Journal of Nonverbal Behav- Zaccaro, S. J., Gilbert, J. A., Thor, K. K., & Mum-
ior, 12, 87–97. ford, M. D. (1991). Leadership and social intelli-
Waldman, D. A., Bass, B. M., & Yammarino, F. J. gence: Linking social perceptiveness to behavioral
(1990). Adding to contingent reward behavior: flexibility. Leadership Quarterly, 2, 317–347.
The augmenting effect of charismatic leadership. Zaccaro, S. J., Mumford, M. D., Connelly, M. S.,
Group & Organization Studies, 15, 381–394. Marks, M. A., & Gilbert, J. A. (2000). Assessment
Wentworth, D. K., & Anderson, L. R. (1984). Emer- of leader problem-solving capabilities. Leadership
gent leadership as a function of sex and task type. Quarterly, 11, 37– 64.
Sex Roles, 11, 513–524.
Zaccaro, S. J. (2002). Organizational leadership and
social intelligence. In R. E. Riggio, S. E. Murphy, Received August 9, 2002
& F. J. Pirozzolo (Eds.), Multiple intelligences and Revision received October 28, 2002
leadership (pp. 29 –54). Mahwah, NJ: Erlbaum. Accepted January 27, 2003 䡲