Journal of Management
The online version of this article can be found at:
2005 31: 126Journal of Management
and Dwight D. Frink
Gerald R. Ferris, Darren C. Treadway, Robert W. Kolodinsky, Wayne A. Hochwarter, Charles J. Kacmar, Ceasar Douglas
Development and Validation of the Political Skill Inventory
On behalf of:
Southern Management Association
can be found at:Journal of ManagementAdditional services and information for
What is This?
- Jan 21, 2005Version of Record >>
by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from by guest on October 11, 2013jom.sagepub.comDownloaded from
10.1177/0149206304271386ARTICLEJournal of Management / February 2005Ferris et al. / Political Skill In ventory
Development and Validation of
the Political Skill Inventory
Gerald R. Ferris*
Department of Management, Florida State University, Tallahassee, FL 32306-1110
Darren C. Treadway
School of Business Administration, University of Mississippi, University, MS 38677-1848
Robert W. Kolodinsky
Department of Management, James Madison University, Harrisonburg, VA 22807-0205
Wayne A. Hochwarter
College of Business, Florida State University, Tallahassee, FL 32306-1110
Charles J. Kacmar
Department of Information Systems, Statistics, and Management Science, College of Business,
University of Alabama, Tuscaloosa, AL 35487-0226
College of Business, Florida State University, Tallahassee, FL 32306-1110
Dwight D. Frink
School of Business Administration, University of Mississippi, University, MS 38677-1848
The present research was developed to examine the conceptualization and measurement of the political
skill construct and to provide validation evidence for the Political Skill Inventory (PSI). The results of three
investigations, involving seven samples, are reported that demonstrate consistency of the factor structure
across studies, construct validity, and criterion-related validity of the PSI. As hypothesized, political skill
was positively related to self-monitoring, political savvy, and emotional intelligence; negatively related to
trait anxiety; and not correlated with general mental ability. Also, the PSI predicted performance ratings of
managers in two samples. Implications of these findings and directions for future research are provided.
Keywords: political skill; savvy; social effectiveness; scale development and validation
†This research was supported in part by the Florida State University Foundation, the PMB and William King Self Endowment
Fund, and by a grant from the Office of Naval Research (ONR N00014-00-1-0595-AA).
*Corresponding author. Tel.: 850 644-3548; fax: 850 644-7843.
E-mail address: firstname.lastname@example.org
Journal of Management, Vol. 31 No. 1, February 2005 126-152
© 2005 Southern Management Associations
For years, scholars and practitioners alike have acknowledged the existence and importance of poli
tics in organizations. Indeed, theory, research, and practice all have considered the types of strategies
and tactics people employ in efforts to behave politically. What we know less about are the characteris
tics that enable one to exercise influence in ways that lead to success. Some have referred to such quali
ties as interpersonal style, “savvy,” “street smarts,” and “political skill.” However, to date, there has
been little effort to move beyond conceptualization to instrument development. Ferris et al. (1999) pro
vided an initial effort when they reported on the development of a concise, unidimensional measure of
political skill. Their work helped establish some support for the construct, but it called for more com
prehensive attempts to fully explore the content domain of political skill and consider its potential
The purpose of the present research is to report the results of three studies designed to develop a
multidimensional Political Skill Inventory (PSI), with item content that more broadly and representa
tively samples from the full domain of the construct. Furthermore, this research offers a more fully
developed conceptualization regarding the dimensions underlying this construct, confirmatory valida
tion of this factorial structure, and evidence of convergent, discriminant, and criterion-related validity.
Political Skill in Organizations
Overview. A perspective shared by many academicians is that organizations are inherently political
arenas (Mintzberg, 1985). In this regard, it is assumed that although performance, effectiveness, and
career success are determined in part by intelligence and hard work, other factors such as social astute-
ness, positioning, and savvy also play important roles (e.g., Luthans, Hodgetts, & Rosenkrantz, 1988;
Mintzberg, 1983). As one of the first to use the term political skill in the scholarly literature, Pfeffer
(1981) argued for a political perspective on organizations. He suggested that political skill is needed to
be successful, and he called for research that would develop a more informed understanding of the
construct. Mintzberg (1983) suggested that political skill referred to the exercise of influence through
persuasion, manipulation, and negotiation.
Although considerable research has examined organizational politics, a serious omission has been
the failure to evaluate the political skill of the influencer, leaving us ill-informed about why influence
efforts are (or are not) successful. Indeed, theory and research largely have assumed that the mere dem
onstration of an influence attempt is synonymous with its effectiveness. However, it is not enough to
study the particular influence tactics or political behaviors that reflect the what of influence. We also
need to critically examine the political skill of the influencer in order to understand the how of influ
ence, which addresses the selection of the most situationally appropriate influence tactics and their
successful execution (Ferris, Hochwarter, Douglas, Blass, Kolodinsky, & Treadway, 2002). In
response to this gap, Ferris et al. (1999) made an initial effort to measure political skill by developing a
concise, six-item, unidimensional scale, with acceptable psychometric properties, for purposes of
preliminary construct exploration.
Definition and specification of the construct domain. In an effort to capture the essential nature of
the construct as Ferris et al. (1999), Mintzberg (1983), and others discussed, we define political skill as
“the ability to effectively understand others at work, and to use such knowledge to influence others to
act in ways that enhance one’s personal and/or organizational objectives” (Ahearn, Ferris,
Hochwarter, Douglas, & Ammeter, 2004: 311). As such, politically skilled individuals combine social
astuteness with the capacity to adjust their behavior to different and changing situational demands in a
Ferris et al. / Political Skill Inventory 127
manner that appears to be sincere, inspires support and trust, and effectively influences and controls
the responses of others.
Politically skilled individuals convey a sense of personal security and calm self-confidence that
attracts others and gives them a feeling of comfort. This self-confidence never goes too far so as to be
perceived as arrogance but is always properly measured to be a positive attribute. Therefore, although
self-confident, those high in political skill are not self-absorbed (although they are self-aware) because
their focus is outward toward others, not inward and self-centered. This allows politically skilled indi
viduals to maintain proper balance and perspective, and also, along with their tendency to be conscien
tious, to ensure that they keep a healthy gauge on their accountability to both others and themselves.
We suggest that people high in political skill not only know precisely what to do in different social
situations at work but how to do it in a manner that disguises any ulterior, self-serving motives and
appears to be sincere. Furthermore, as we discuss later in this article, we see political skill as independ
ent from general mental ability and related to personality traits and other interpersonally oriented con
structs such as self-monitoring and emotional intelligence, but not too highly or to the degree indicat
ing construct redundancy. In terms of its derivation, we borrow from others who have suggested the
usefulness of taking an integrative dispositional-situational approach to personality (e.g., Murtha,
Kanfer, & Ackerman, 1996) and social effectiveness (e.g., Buck, 1991). Specifically, we believe there
are aspects of political skill that are dispositional, but we see other aspects that can be developed or
shaped through a combination of formal and informal developmental experiences (e.g., Ferris,
Anthony, Kolodinsky, Gilmore, & Harvey, 2002).
Content Validity and Dimensionality of Political Skill
Need and rationale. Careful examination of the organizational politics literature, with particular
reference to that segment relating to political skill (i.e., even if not explicitly referred to by that term),
indicates several important aspects that should be included in any representative measure of the politi-
cal skill construct. These aspects or dimensions are social astuteness, interpersonal influence, net-
working ability, and apparent sincerity. Certainly, the ability to read and understand people, and being
able to act on that knowledge in influential ways (i.e., consistent with our definition of political skill)
represent two important dimensions of the construct reflecting social astuteness and interpersonal
influence (e.g., Mintzberg, 1983; Pfeffer, 1981; Snyder, 1987). In addition, there are other aspects of
political skill that are very important, according to power and politics scholars, and therefore need to
be represented in a content-valid representation of the political skill construct.
Connections, friendships, network building, alliances, and coalition building are critical for indi
viduals to navigate the politics of organizations and thus represent an important aspect of political skill
(e.g., Bacharach & Lawler, 1998; Pfeffer, 1981, 1992). Bacharach and Lawler concluded, “To pursue
political action, it is inevitable that actors in the organization align themselves with others” (1998: 85).
Furthermore, Pfeffer argued, “Having connections, having allies, is important for developing and
exercising influence” (1992: 175). In addition, Luthans et al. defined networking as “a system of inter
connected or cooperating individuals. It is closely associated with the dynamics of power and the use
of social and political skills” (1988: 119-120). Furthermore, these activities associated with network
ing were, by far, the most dominant activities exhibited by successful managers.
A final aspect of political skill represented in the organizational politics literature is genuineness or
sincerity, and this reflects the true execution component of political skill. It is not just what behaviors
individuals exhibit but, more so, that they demonstrate influence attempts in ways that are intended to
appear to be sincere and genuine, devoid of ulterior motive, and that inspire trust and confidence. Jones
(1990) appealed for research examining the way influence attempts were executed, focusing on the
interpersonal style component that he argued gave the appearance of sincerity.
128 Journal of Management / February 2005
The forgoing discussion reflects the four critical dimensions of political skill that should be
included in any representative, content-valid measure of the political skill construct. The Ferris et al.
(1999) six-item, unidimensional scale attempted to address the understanding or social astuteness
dimension, as well as the interpersonal influence component (i.e., even though these two dimensions
did not separate into distinguishable factors). However, the Ferris et al. scale is deficient by neglecting
to address the networking ability and apparent sincerity dimensions.
Therefore, we suggest the need for a multidimensional, content-valid measure of political skill that
identifies and assesses the four key dimensions reflected in the organizational politics literature. Each
dimension is described in detail below.
Social astuteness. Individuals possessing political skill are astute observers of others and are keenly
attuned to diverse social situations. They comprehend social interactions and accurately interpret their
behavior, as well as that of others, in social settings. They have strong powers of discernment and high
self-awareness. Pfeffer referred to this characteristic as “sensitivity to others,” and he argued, “Some
what ironically, it is this capacity to identify with others that is actually critical in obtaining things for
oneself” (1992: 173). Socially astute individuals often are seen as ingenious, even clever, in dealing
with others. As such, this dimension of political skill would be expected to relate most strongly (of all
the dimensions) to supervisor evaluations of an employee’s job performance because of the
employee’s social astuteness at presenting his or her work behavior in the best possible light. In sum
mary, people high in social astuteness have an accurate understanding of social situations as well as the
interpersonal interactions that take place in these settings.
Interpersonal influence. Politically skilled individuals have a subtle and convincing personal style
that exerts a powerful influence on those around them. Individuals high on interpersonal influence
nonetheless are capable of appropriately adapting and calibrating their behavior to each situation in
order to elicit particular responses from others. Indeed, aspects of the interpersonal influence dimen-
sion capture what Pfeffer (1992) referred to as “flexibility,” which involves adapting one’s behavior
situationally to different targets of influence in different contextual conditions in order to achieve one’s
Networking ability. Individuals with strong political skill are adept at developing and using diverse
networks of people. People in these networks tend to hold assets seen as valuable and necessary for
successful personal and organizational functioning. By the sheer force of their typically subtle style,
politically skilled individuals easily develop friendships and build strong, beneficial alliances and
coalitions. Furthermore, because social networks are deliberately constructed structures, individuals
high in networking ability ensure they are well positioned in order to both create and take advantage of
opportunities (Pfeffer, 1992). Masters of the quid pro quo, they are often highly skilled negotiators and
deal makers, and adept at conflict management.
Apparent sincerity. Politically skilled individuals appear to others as possessing high levels of
integrity, authenticity, sincerity, and genuineness. They are, or appear to be, honest, open, and forth
right. This dimension of political skill strikes at the very heart of whether influence attempts will be
successful because it focuses on the perceived intentions (i.e., as assessed by the target of influence) of
the behavior exhibited (i.e., by the actor). Indeed, perceived intentions or motives are important and
have been argued to alter the interpretation and labeling of behavior. As noted by Jones (1990), influ
ence attempts will be successful only when actors are perceived as possessing no ulterior motives.
Because their actions are not interpreted as manipulative or coercive, individuals high in apparent sin
cerity inspire trust and confidence in and from those around them.
Ferris et al. / Political Skill Inventory 129
Convergent and Discriminant Validity
Convergent and discriminant validity, respectively, reflect the extent to which a measure relates to
similar constructs and does not relate to constructs from which it should differ. In this section, hypoth
eses are formulated concerning the constructs to which political skill (i.e., total score and separate
dimensions) should and should not be related.
Personality and social effectiveness constructs. We conceptualize the political skill construct as
overlapping somewhat with selected personality traits and with other related constructs that purport to
measure social effectiveness (Ferris, Perrewé, & Douglas, 2002). However, we also believe that politi
cal skill reflects its distinctiveness as a construct that is sufficiently different from others. Therefore,
we expect that political skill would be related positively and perhaps even significantly with constructs
such as self-monitoring and conscientiousness. However, these relationships should not be so great in
magnitude as to indicate construct redundancy.
Other social effectiveness constructs, which have been characterized in ways that suggest consider
able overlap with political skill, are political savvy and emotional intelligence. Political savvy sug
gests adeptness at the nuances of politics in organizations and, as such, should be related positively,
and perhaps significantly, to the composite measure of political skill. Work by Chao, O’Leary-Kelly,
Wolf, Klein, and Gardner (1994) indicated that political savvy might be driven by a knowledge or
understanding component but could include an implicit capacity to act on that knowledge, although
that aspect is not made clear in its derivation.
Emotional intelligence has received considerable attention in the popular, business, and research
press in the past 8 years, primarily as a function of Goleman’s (1995, 1998) best-selling books. Such
discussions of emotional intelligence are broad and could be construed as being redundant with politi-
cal skill. However, more careful reading of the scientific research indicates that emotional intelligence
focuses predominantly on the emotion-based aspects of interpersonal effectiveness, influence, and
control. Conversely, we see political skill as incorporating knowledge and skill that go beyond emo-
tions. Therefore, particularly because of Goleman’s broad characterization of emotional intelligence
(i.e., which has led some to suggest that Goleman regards emotional intelligence as including every-
thing except general mental ability or intelligence; Hedlund & Sternberg, 2000) and the foregoing dis
cussion of personality and social effectiveness constructs, we hypothesize the following:
Hypothesis 1: The political skill total score will correlate significantly and positively with self-monitoring,
conscientiousness, political savvy, and emotional intelligence.
Of the four political skill dimensions, social astuteness should be most strongly related to self-
monitoring, conscientiousness, and political savvy. The social astuteness dimension best captures the
essence of the self-monitoring construct, as explained by Snyder (1987). Furthermore, the attention to
detail, and, as Pfeffer stated, the “almost clinical interest in the observation of behavior” (1992: 173),
relates this dimension of political skill most strongly with conscientiousness.
Notions of “savvy” in general, and “political savvy” in particular, make reference to a degree of
understanding that is closely related to the social astuteness dimension. Chao et al. (1994) identified a
dimension of organizational socialization they called “politics.” Closer inspection of the item content
of this dimension reveals that it is actually measuring political savvy or understanding. Items such as “I
have learned how things ‘really work’ on the inside of this organization,” “I know who the most influ
ential people are in my organization,” and “I have a good understanding of the motives behind the
actions of other people in the organization” highlight the importance of the savvy or understanding
aspect of politics, thus suggesting its relationship to the social astuteness PSI dimension. Therefore, we
formulate the following hypothesis:
130 Journal of Management / February 2005
Hypothesis 2: Social astuteness will demonstrate the strongest positive correlations with self-monitoring, con
scientiousness, and political savvy of any of the PSI dimensions.
Influence tactics. In addition, we believe that political skill reflects the capacity to effectively exer
cise influence over others at work. Therefore, political skill should be related to particular types of
influence tactics, as Kipnis, Schmidt, and Wilkinson (1980) discussed, such as upward appeal and
coalition, but less so to assertiveness. Kipnis et al. discussed the upward appeal tactic as involving
obtaining the support of individuals higher up in the organizational hierarchy. Coalition tactics refer to
mounting coworker or subordinate support to reinforce a position taken or resources requested, count
ing on a strength-in-numbers approach. Finally, assertiveness involves demanding, ordering, setting
deadlines and checking up on others in order to exercise influence. Indeed, we suggest that when those
high in political skill engage in influence tactics, they do so in an effective way. However, those high in
political skill might simply decide to avoid some influence tactics in favor of others.
Hypothesis 3: The political skill total score will correlate significantly and positively with upward appeal and
coalition influence tactics, but nonsignificantly with assertiveness.
Concerning the dimensions of political skill, we argue that a significant positive relationship
between the coalition influence tactic and the networking ability dimension of political skill will sur
face. We make a similar argument for the upward appeals influence tactic because networks and con-
nections are necessary to the facilitation of upward appeal because such linkages allow one to operate
from a position of greater strength (i.e., through such connections and social capital). Similarly, net-
working ability should be most strongly related to the use of assertiveness as an influence tactic. The
use of assertiveness, as a way to influence others, can be intricate, and its effects sometimes can be pos-
itive and sometimes negative. We suggest that individuals who possess a high degree of networking
ability will be well positioned (e.g., by virtue of the connections, alliances, and social capital they
control) to employ assertiveness as a successful influence tactic.
Hypothesis 4: Networking ability will demonstrate the strongest positive correlation with the upward appeal,
coalitions, and assertiveness tactics of any of the PSI dimensions.
Trait anxiety. Furthermore, we suggest that political skill demonstrates an inverse relationship with
trait anxiety, which reflects “relatively stable individual differences in anxiety-proneness, that is, to
differences between people in the tendency to perceive stressful situations as dangerous or threatening
and respond to such situations with elevations in the intensity of their state anxiety (S-Anxiety) reac
tions” (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983: 4). We believe political skill generates
an increased sense of self-confidence and personal security because such individuals should experi
ence a greater degree of control over activities that transpire in social interactions at work.
Indeed, Perrewé, Ferris, Frink, and Anthony (2000) argued that such greater self-confidence and
control should lead individuals high in political skill to experience significantly less stress or anxiety at
work. Furthermore, Perrewé, Zellars, Ferris, Rossi, Kacmar, and Ralston (2004) recently reported
results supportive of this notion, demonstrating that political skill neutralized the dysfunctional effects
of role conflict on strain for four of five behavioral and physiological strain measures (i.e., even pro
ducing a reduction in systolic blood pressure). Consequently, political skill may serve as an antidote, of
sorts, to the negative consequences of stress.
Hypothesis 5: The political skill total score will demonstrate a significant negative correlation with trait
Ferris et al. / Political Skill Inventory 131
Concerning the dimensions of political skill, it appears that interpersonal influence should exhibit
the strongest negative relationship with trait anxiety. The heightened sense of personal security and
self-confidence are likely associated with perceiving greater control over one’s work environment,
which includes the people with whom one interacts. Such feelings are likely reflective of the percep
tions of greater interpersonal control they derive from past experiential success at exercising interper
sonal influence. Therefore, we argue that feelings of control and personal security that result from
being high on the interpersonal influence dimension of political skill would provide a comfort level
that would result in reduced anxiety, thus exhibiting the strongest negative relationship of any of the
Hypothesis 6: Interpersonal influence will demonstrate the strongest negative correlation with trait anxiety of
any of the PSI dimensions.
General mental ability. The principal argument for the discriminant validity of political skill relates
to the expected relationship with general mental ability (GMA). Typically, scholars bear the burden of
demonstrating empirically that their interpersonally oriented constructs are not simply a small part of,
and thus not distinguishable from, GMA. Supporters of the “GMA-dominant” position (e.g.,
Gottfredson, 1997) would argue that political skill (and other measures of social effectiveness) could
be effectively subsumed under GMA, thus implying that political skill would reflect a large correlation
with GMA. However, we see political skill as independent of GMA, and Ferris et al. (1999) reported a
nonsignificant relationship between their six-item measure of political skill and GMA (i.e., r = –.08,
n.s.), thus demonstrating support for its discriminant validity.
Hypothesis 7: The political skill total score, and each of the PSI dimensions, will demonstrate zero correlations
Although there has been limited previous work concerning the criterion-related validity, we pro-
pose that political skill will demonstrate significant predictive ability, particularly for work perfor
mance criteria that are subjectively assessed by others (e.g., supervisors, peers, etc.). Ahearn et al.
(2004) found leader political skill to be a significant predictor of team performance. Furthermore, Hig
gins (2000) reported that political skill was related significantly to recruitment interviewer ratings and
evaluations of job applicants. These two studies used the Ferris et al. (1999) measure of political skill.
Hypothesis 8: The political skill total score will demonstrate significant positive prediction of work perfor
Finally, regarding criterion-related validity, social astuteness would be expected to relate most
strongly to supervisor evaluations of job performance because of the employee’s adeptness at present
ing his or her work behavior in the best possible light. Socially astute individuals use their ability to
read situations and people and use their flexibility as input to the design of tailored attempts to
Hypothesis 9: Social astuteness will demonstrate the strongest positive prediction of job performance ratings
of any of the PSI dimensions.
132 Journal of Management / February 2005
Plan of Research and Sequence of Studies
As an expansion of Ferris et al.’s (1999) earlier work, this research involves three studies and seven
samples that are designed to examine, as comprehensively as possible, the psychometric properties of
the PSI. Study 1 focuses on creating the PSI, assessing its dimensionality, and providing initial evi
dence of convergent and discriminant validity. Study 2 attempts to confirm the factor structure and
construct dimensionality (Sample 1) and to provide additional evidence of convergent and
discriminant validity, including additional measures of political savvy, emotional intelligence, and
GMA (Samples 2 and 3). Finally, Study 3 attempts to demonstrate the criterion-related validity of the
PSI by assessing its capacity to predict subordinate evaluations of leadership effectiveness (Sample 1)
and supervisor ratings of subordinate performance (Sample 2).
Study 1: PSI Development and Initial Validation
Two samples were obtained in an effort to further explore the political skill construct. First, a sam
ple of 226 undergraduate students at a large southern university completed surveys during class time.
All students present in class that day participated in the data collection. The average age of the respon-
dents in this sample was 22.61 (σ = 6.39), 50% of the sample was female, and the majority were part-
time workers. For the second sample, the Human Resources department at a large university in the
southeastern United States generated a list of 220 employees classified as either “managerial or
Surveys were distributed and subsequently returned to the researchers via interoffice mail. A total
of 124 employees completed surveys for a response rate of 56%. The average age of the respondents in
the second sample was 39.54 (σ = 9.74), nearly 70% were female, and the average organizational ten-
ure was 9.47 years (σ = 7.07). In scale development particularly, larger samples are necessary to help
ensure a greater degree of stability of the obtained results, so we combined the two data sets (N = 350).
Political skill item pool. A total of 40 items were generated to representatively assess the political
skill construct and specifically to reflect the four dimensions believed to comprise political skill, thus
maximizing content validity. In so doing, we built on, and expanded, the earlier work on the measure
ment of political skill by Ferris et al. (1999). We included the original six items developed by Ferris
et al. in the total item pool and conducted a systematic review of the literature on organizational poli
tics, with particular reference to skills in using politics effectively. In so doing, the authors wrote items
to representatively reflect the key areas identified to best characterize the political skill construct and,
as such, employed a “logical partitioning” approach to deductive scale development discussed by
Hinkin (1995). Respondents were asked to indicate the extent to which they agreed with each state
ment about themselves at work, using a 1 (strongly disagree)to7(strongly agree) Likert-type scale. A
copy of the 40 items generated from the above process is included in the appendix.
Self-monitoring. Self-monitoring reflects the extent to which individuals monitor and control the
images they project in social situations, and it was measured with the 18-item instrument presented in
Snyder (1987). The coefficient alpha for this scale was .75.
Ferris et al. / Political Skill Inventory 133
Personality. Conscientiousness was examined using items from McCrae and Costa (1987). Consci
entiousness suggests the tendency to be organized, disciplined, and dependable. The coefficient alpha
for the 14-item Conscientiousness Scale was .87.
Influence tactics. The Kipnis et al. (1980) measure of influence tactics was used to assess the fre
quency with which individuals employed upward influence tactics. Specifically, the scale contained
upward appeal (one-item measure), coalition (three items, α = .63), and assertiveness (six items, α =
Trait anxiety. Trait anxiety identifies tendencies for individuals to naturally be anxiety prone and
perceive stressful situations as threatening, and it was measured with the 20-item Spielberger et al.
(1983) State-Trait Anxiety Inventory (Form Y-2). The coefficient alpha for this scale was .88.
Social desirability. Social desirability was measured using the 10-item scale developed by Strahan
and Gerbasi (1972). This measure had a coefficient alpha of .71.
Item analyses. Because our interest was in scale development, we retained only those items that
provided the best representation of the political skill construct, while measuring this construct in the
most parsimonious way. Thus, we undertook several phases of scale construction. First, we conducted
item analyses, retaining those with the highest item-to-total correlations (i.e., item-to-total correla-
tions of .40 or greater; Nunnally, 1978). This resulted in the elimination of 5 items (i.e., Items 1, 11, 27,
28, and 40 in the appendix). Next, we examined correlations of political skill items with the total score
on the Social Desirability Scale and deleted items with statistically significant correlations at p < .05.
This resulted in the elimination of an additional 10 items (i.e., Items 6, 12, 14, 15, 17, 22, 33, 34, 35,
and 36 in the appendix).
Finally, to examine the extent to which there might be remaining items that demonstrated problem-
atic high cross loadings on factors that precluded clear interpretation, we conducted a preliminary
principal components analysis with oblique rotation. The pattern of factor loadings showed that 7
items (i.e., Items 2, 4, 7, 10, 18, 19, and 20 in the appendix) loaded on two or more factors greater than
.35, with the highest loading not being on the intended factor. Therefore, these 7 items were elimi
nated, resulting in this three-step item-reduction procedure yielding a set of 18 items that met the crite
ria for item retention and would be subjected to principal axis factoring to assess the factor structure of
The item-to-total correlations for these 18 items ranged from .42 to .71. Moreover, the total score
for the 18 items did not correlate significantly with the Social Desirability Scale total score (r = .02,
n.s.). Also, the correlations and mean differences of political skill with gender and age were examined,
and no systematic relationships were detected. The internal consistency reliability estimate for the
resulting 18-item scale total score was .90 (the 18 items making up this scale are asterisked in the
appendix). Also, the original 6 items comprising the Ferris et al. (1999) Political Skill Scale were
included in this 40-item pool used to generate the final 18-item PSI. Of those 6 items, 3 items were
retained as part of the final 18-item inventory (the Ferris et al. 6 items are noted in the appendix).
Dimensionality of political skill. To determine the factor structure of the 18-item PSI, and following
the item-reduction procedures specified above, we performed factor analysis on the 18 items, using the
principal axis method and oblique, direct oblimin factor rotation. It has been noted that oblique factor
rotation generally is more desirable than orthogonal rotation (e.g., Hair, Anderson, & Tatham, 1987) at
134 Journal of Management / February 2005
this early stage of scale development because of the fewer constraints it imposes. Furthermore, oblique
factor rotation is most appropriate when the a priori theory indicates that obtained factors or dimen
sions are likely correlated. A four-factor solution emerged that satisfied the Kaiser-Guttman criterion
of retaining only those factors with eigenvalues exceeding 1.0. Fabrigar, Wegener, MacCallum, and
Strahan (1999) recommended using relevant theory and multiple methods in factor retention decisions
in order to balance the need for parsimony with that of plausibility.
As shown in Table 1, eigenvalues ranged from 1.19 to 6.98, with 63% of the total variance
explained. The factor explaining most of the variance was networking ability (six items, indicated by
italics in the table), with 39% of the variance explained. Factor 2, apparent sincerity (three items, itali
cized) explained 10% of the variance, social astuteness (five items, italicized) accounted for 7%, and
interpersonal influence (four items, italicized) explained 6.6% of the total variance.
Reliability and factor correlations. As noted in Table 1, internal consistency indices revealed reli
ability estimates ranging from .78 (interpersonal influence) to .87 (networking ability) for dimensions
contained within the modified 18-item scale. These values are above the .70 level recommended by
Nunnally (1978). The lower diagonal of Table 2 reports the intercorrelations of the political skill
dimensions, which range in magnitude from .36 to .57, reflecting only modest relationships.
Fit statistics and alternative models. Confirmatory factor analysis was conducted to test the
hypothesis that four dimensions comprised the political skill construct. Several recommended mea-
sures of overall goodness of fit were used, including the Comparative Fit Index (CFI), the Normed Fit
Index (NFI), the Tucker-Lewis Index (TLI) (Tucker & Lewis, 1973) or Non-Normed Fit Index
(NNFI), the Goodness-of-Fit Index (GFI), the Adjusted Goodness-of-Fit Index (AGFI), the root mean
square error of approximation (RMSEA), the standardized root mean square residual (SRMR), and the
ratio of chi-square relative to the degrees of freedom (χ
/df) (e.g., Hu & Bentler, 1999; La Du &
Tanaka, 1989; Wheaton, Muthen, Alwin, & Summers, 1977).
Values of .90 or higher are desired and presumed to result in an acceptable model fit to the data for
the CFI, NFI, NNFI, and GFI (e.g., Hatcher, 1994; Medsker, Williams, & Holohan, 1994; Mulaik,
James, Van Alstine, Bennet, Lind, & Stillwell, 1989), whereas a value higher than .80 is recommended
for the AGFI (Gefen, Straub, & Boudreau, 2000). The appropriate level for the RMSEA was estab-
lished at “a cutoff value close to .06” (Hu & Bentler, 1999: 27), and for the SRMR, “a cutoff value close
to .08” (Hu & Bentler, 1999: 27). Values of less than 5 for the χ
/df ratio indicate acceptable model fit
(Wheaton et al., 1977).
The four-factor proposed model was tested for fit and compared to the fit indices for one-, two-, and
three-factor models (Anderson & Gerbing, 1988). The results from structural equation modeling anal
yses (LISREL 8; Jöreskog & Sörbom, 1993) are presented in Table 3 and reveal that the four-factor
solution was superior to the other three models. Specifically, the one-factor model results indicated an
unacceptable fit, whereas the four-factor solution results were at (or just below) acceptable levels, indi
cating a better fit relative to the one-factor solution. Based on these findings, and the chi-square differ
ence test between the one- and four-factor models (χ
difference = 454.64, df = 6), the unidimensional
solution was rejected in favor of the multidimensional model.
Both two-factor and three-factor models were also tested and compared with the four-factor target
model. The most plausible two-factor model was tested that had the social astuteness and networking
ability items comprising one factor and the interpersonal influence and apparent sincerity items mak
ing up a second factor. The two-factor model indicated an unacceptable fit and, along with the signifi
cant chi-square difference test (χ
difference = 256.92, df = 5), indicates that the four-factor model is
superior to the two-factor model. In addition, two other two-factor models were run, and in both cases
the fit statistics were unacceptable, and the chi-square difference tests demonstrated the superiority of
the four-factor model.
Ferris et al. / Political Skill Inventory 135
Factor Analysis and Item Analysis Results (Study 1)
Factor 1 Factor 2 Factor 3 Factor 4
Networking Apparent Social Interpersonal Item-Total Desirability
Item Ability Sincerity Astuteness Influence rr
1. I spend a lot of time and effort at work networking with others. .79 .01 .08 .03 .59 .08
2. At work, I know a lot of important people and am well connected. .78 .01 .08 .04 .71 .08
3. I am good at using my connections and networks to make things happen at work. .73 .01 .13 .06 .63 .04
4. I have developed a large network of colleagues and associates at work who
I can call on for support when I really need to get things done. .67 .26 .03 .07 .58 –.02
5. I spend a lot of time at work developing connections with others. .60 .07 .01 .13 .53 .01
6. I am good at building relationships with influential people at work. .47 .16 .25 .17 .60 .05
7. It is important that people believe I am sincere in what I say and do. .12 .72 .03 .14 .51 –.03
8. When communicating with others, I try to be genuine in what I say and do. .03 .70 .09 .01 .42 –.08
9. I try to show a genuine interest in other people. .02 .58 .12 .24 .53 .01
10. I always seem to instinctively know the right thing to say or do to influence others. .16 .21 .66 .14 .59 .01
11. I have good intuition or savvy about how to present myself to others. .15 .03 .64 .13 .68 –.06
12. I am particularly good at sensing the motivations and hidden agendas of others. .04 .17 .63 .16 .45 .03
13. I pay close attention to people’s facial expressions. .09 .15 .45 .20 .49 .03
14. I understand people very well. .07 .11 .40 .18 .53 .04
15. It is easy for me to develop good rapport with most people. .08 .09 .11 .72 .55 –.06
16. I am able to make most people feel comfortable and at ease around me. .02 .01 .07 .72 .55 –.01
17. I am able to communicate easily and effectively with others. .08 .04 .11 .56 .57 .00
18. I am good at getting people to like me. .07 .11 .11 .46 .53 .04
Eigenvalue 6.98 1.86 1.29 1.19
Percentage of variance explained 39.00 10.35 7.17 6.63
Cumulative percentage of variance explained 39.00 49.14 56.31 62.94
Coefficient alpha reliability estimates .87 .81 .79 .78
Correlations of Political Skill Dimensions
Networking Apparent Social Interpersonal
Ability Sincerity Astuteness Influence
Networking ability 4.90 1.04 .87 3.51 .66 .82 — .28 .56 .49
Apparent sincerity 5.92 0.84 .81 4.05 .56 .73 .36 — .47 .57
Social astuteness 5.18 0.91 .79 3.72 .58 .79 .57 .45 — .69
Interpersonal influence 5.60 0.87 .78 4.01 .54 .77 .53 .50 .56 —
Note: Coefficients below the diagonal reflect the factor correlations for Study1(N = 350), and those above the diagonal reflect the factor correlations for Study 2 (N = 93). All correlations in the
table are significant p < .001.
a. Indicates Study 1 results.
b. Indicates Study 2 results.
Structural Equation Modeling Statistics for One-, Two-, Three-, and Four-Factor Models in Studies 1 and 2
Study 1 Models Study 2 Models
Fit Indices 4-Factor 3-Factor 2-Factor 1-Factor 4-Factor 3-Factor 2-Factor 1-Factor
Comparative Fit Index (CFI) .92 .87 .86 .80 .94 .91 .91 .91
Normed Fit Index (NFI) .88 .84 .82 .77 .87 .84 .85 .84
Non-Normed Fit Index (NNFI) (or Tucker-Lewis Index) .90 .85 .83 .77 .92 .88 .89 .88
Goodness-of-Fit Index (GFI) .91 .87 .84 .80 .89 .87 .88 .87
Adjusted Goodness-of-Fit-Index (AGFI) .87 .82 .79 .74 .84 .81 .82 .81
Root mean square error of approximation (RMSEA) .07 .09 .10 .12 .06 .08 .07 .08
Standardized root mean square residual (SRMR) .05 .08 .07 .08 .06 .07 .07 .07
Ratio of chi-square to degrees of freedom (χ
/df) 2.61 3.67 4.50 5.98 1.75 2.11 2.00 2.08
The most plausible three-factor model also was tested where the three factors reflected the network
ing ability, social astuteness, and interpersonal influence dimensions, with the three items from the
apparent sincerity dimension spread across these three factors. The fit statistics for the three-factor
model, and the significant chi-square difference test (χ
difference = 142.80, df = 3), indicate that the
four-factor model also provides a better fit than the three-factor model. An additional three-factor
model was run, and it too was found to be inferior to the four-factor model according to the fit statistics
and chi-square difference test.
Convergent and discriminant validity. As argued, we expect political skill to be positively corre
lated with self-monitoring and conscientiousness (Hypothesis 1). Also, we need to demonstrate that
political skill is not simply synonymous with influence tactics of upward appeal and coalitions, and
that it is not related to assertiveness (Hypothesis 3). Finally, we believe that political skill should be
inversely related to trait anxiety such that politically skilled individuals are apt to experience less strain
The correlations of the PSI composite measure, the four dimensions of political skill, and the other
scales are presented in Table 4. Preliminary results provide evidence for the convergent and
discriminant validity of the political skill construct. As expected, the overall political skill construct
was significantly and positively related to self-monitoring (r = .39, p < .001) and conscientiousness
(r = .31, p < .001), thus supporting Hypothesis 1.
Hypothesis 2 proposed that the social astuteness dimension of political skill would demonstrate the
strongest relationship with both self-monitoring and conscientiousness. For self-monitoring (r = .37,
p < .001) and conscientiousness (r = .27, p < .001), social astuteness did exhibit the highest correla-
tions of the four dimensions. However, the differences between the correlations for these two con-
structs with the other three PSI dimensions were not statistically significant when conducting signifi-
cance tests for the differences between dependent correlation coefficients (e.g., Cohen & Cohen,
1983; because these are directional hypotheses, the statistical tests for differences use one-tailed sig-
nificance tests, with N – 3 degrees of freedom). Thus, Hypothesis 2 failed to receive full support.
Results also showed evidence that political skill is not redundant with influence tactics. Indeed,
none of the correlations between political skill and the individual influence tactics approached a level
to suggest multicollinearity (r > .80, Lewis-Beck, 1980). Specifically, political skill was related to the
upward appeal (r = .25, p < .001) and coalition (r = .21, p < .001) influence tactics. Furthermore, it was
suggested that political skill, being more indirect, would not likely be correlated with assertiveness.
Indeed, the correlation of political skill with assertiveness was not significant (r = .09, n.s.). These
results for the relationships with influence tactics provide support for Hypothesis 3.
Because the nature of networking involves involvement with others and being able to mobilize
coalitions for influence, networking ability was hypothesized to demonstrate the strongest positive
relationship with the upward appeal and the coalition influence tactics, and that was found to be the
case (r = .30, p < .001; r = .31, p < .001, respectively). Significance tests for the difference between cor
relations reported that the networking ability–upward appeal correlation was significantly larger than
the correlations of any of the other PSI dimensions with upward appeal (i.e., t = 1.89, p < .05 for social
astuteness; t = 5.29, p < .001 for interpersonal influence; and t = 2.92, p < .01 for apparent sincerity).
Finally, it was hypothesized that the networking ability dimension would provide the positioning to
allow one to effectively use assertiveness. Indeed, networking ability demonstrated a significant posi
tive relationship with assertiveness (r = .18, p < .001), which was found to be significantly greater than
the correlations of any of the other PSI dimensions with assertiveness (i.e., t = 1.43, p < .10 for social
astuteness; t = 4.98, p < .001 for interpersonal influence; and t = 3.88, p < .001 for apparent sincerity).
These results lend support for Hypothesis 4.
Political skill exhibited a significant negative correlation with trait anxiety (r = .31, p < .001), pro
viding support for Hypothesis 5. It was hypothesized that feelings of control and personal security,
138 Journal of Management / February 2005
Descriptive Statistics, Correlations of Political Skill With Other Measures, and Competitive Prediction of
Other Measures From Political Skill Dimensions and Ferris et al. (1999) Measure (Study 1)
Political Social Interpersonal Networking Apparent Ferris et al.,
Measure MSDα Skill Total Astuteness Influence Ability Sincerity 6-item Measure
Self-monitoring 1.46 0.21 .75 .39*** .37*** .30*** .37*** .11 .30***
(.21**) (.12) (.23***) (–.13*) (.01)
Conscientiousness 3.68 0.63 .87 .31*** .27*** .25*** .24*** .24*** .27***
(.09) (.02) (.08) (.10) (.12)
Upward appeal 2.66 1.11 — .25*** .21*** .04 .30*** .13* .10
(.14) (–.23*) (.33***) (.08) (–.03)
Coalition 2.55 0.83 .63 .21*** .14** .04 .31*** .03 .12*
(.02) (–.18*) (.40***) (–.05) (.02)
Assertiveness 2.33 0.78 .79 .09 .11* –.07 .18*** –.05 –.01
(.15*) (–.20*) (.26***) (–.07) (–.07)
Trait anxiety 1.99 0.46 .88 –.31*** –.28*** .37*** .20*** –.15** –.34***
(–.11) (–.28***) (.04) (.08) (–.10)
M 5.29 5.14 5.60 4.90 5.94 5.48
SD 0.74 0.89 0.84 1.02 0.84 0.73
α .90 .79 .78 .87 .81 .73
Note: Due to missing values, the sample size ranges from 326 to 350. The six-item Ferris et al. (1999) Political Skill Scale correlated r = .78*** with the Political Skill Inventory (PSI) total score,
r = .66*** with the social astuteness dimension, r = .81*** with the interpersonal influence dimension, r = .54*** with the networking ability dimension, and r = .52*** with the apparent sincer
ity dimension. The values in parentheses are the standardized regression coefficients for the prediction of each of the constructs in the left column from the four political skill dimensions and the
Ferris et al. (1999) six-item scale. No reliability coefficient is reported for upward appeal because it is a one-item measure.
which result from scoring high on the interpersonal influence dimension of political skill, would be
associated with reduced anxiety, thus exhibiting the strongest negative relationship of any of the politi
cal skill dimensions. Results supported this proposal, showing that interpersonal influence reflected
the highest correlation with trait anxiety of any of the four dimensions (r = .37, p < .001). Furthermore,
when conducting significance tests between correlations of interpersonal influence and trait anxiety
with each other PSI dimension and trait anxiety, all correlations were significantly smaller in magni
tude (i.e., t = 1.93, p < .05 for social astuteness; t = 3.50, p < .001 for network building; and t = 4.40, p <
.001 for apparent sincerity). This lends strong support to Hypothesis 6.
Measurement and Predictive Properties of
18-Item PSI Versus Ferris et al. (1999) Measure
In the development of a new scale, it is important to establish that the new measure demonstrates
psychometric properties that reflect improvements over previous measures. Table 4 reports the means,
standard deviations, reliability estimates, and correlations of the new 18-item PSI in comparison to the
original Ferris et al. (1999) 6-item Political Skill Scale and demonstrates how both measures relate to
other constructs. The relationships of the 6-item scale and the 18-item measure with self-monitoring,
conscientiousness, and trait anxiety are quite similar. However, the 6-item measure shows very little
relationship with the influence tactics.
To assess the predictive effectiveness of the PSI dimensions compared to the Ferris et al. (1999) 6-
item measure, regression equations were computed with the four political skill dimensions and the 6-
item scale entered simultaneously as competitive predictors, and the constructs presented in the left
column of Table 4 as outcomes. The beta coefficients for each of the predictors and their significance
levels appear in parentheses in Table 4. For the prediction of conscientiousness, none of the regression
coefficients for the five predictors achieved significance. For the remaining outcomes, between one
and three of the PSI dimensions demonstrated significant prediction. However, the interesting finding
seen in Table 4 is that the Ferris et al. measure was not a significant predictor in any of the six regres-
sion equations. Overall, the new PSI demonstrates an improvement over the initial Ferris et al. scale.
Study 2: Factor Structure
Confirmation and Construct Validity
This study had two major purposes. The first was to ensure that the factor structure derived in Study
1 was not an artifact of the survey design, data collection method, or sample obtained for analysis.
Therefore, we attempted to replicate the factor structure from Study 1 using a broader range of respon
dents, settings, and survey methods in Sample 1. Specifically, the sample used students to distribute
surveys containing the 18-item PSI to full-time employees.
The second purpose of Study 2 was to both replicate and expand on the convergent and discriminant
validity evidence reported in Study 1, in two different samples (i.e., Samples 2 and 3). In Sample 2,
data were collected on the 18-item PSI, the same personality and social influence constructs gathered
in Study 1, and a political-savvy construct (Chao et al., 1994) in an effort to both replicate results from
the first study and to extend construct validity evidence. Finally, Sample 3 examined the relationships
of political skill with GMA and emotional intelligence.
Sample 1: Participants and Procedures
Students at a large university in the southeastern United States each were given three surveys and
asked to find three full-time employees, working a minimum of 5 years, who would complete the sur
140 Journal of Management / February 2005
veys. Occupations in the sample included patrol officer, human resource manager, chief executive offi
cer, and accountant. The student received course credit for securing appropriate completion of the
The sample consisted of 85 male employees (44%) and 108 female employees (56%), and the aver
age age was approximately 42 years (M = 42.24, σ = 10.75). Respondents had an average of almost 20
years of full-time work (M = 19.55, σ = 10.52), had worked in their current position for roughly 7 years
(M = 7.22, σ = 8.82), and had been with their current organization for almost 9 years (M = 8.86, σ =
8.95). Participants reported supervising an average of roughly 15 employees (M = 14.94, σ = 50.87,
range 0-524). Finally, 72 respondents (37%) reported professional staff as their position followed by
47 (24%) reporting middle management, 42 (22%) reporting nonmanagement, 18 (9%) reporting
upper management, and 12 (6%) reporting other.
Sample 1: Results
The four-factor solution met most of the standards to demonstrate adequate fit of the model,
whereas the one-factor solution had less acceptable fit, as seen in Table 3. A chi-square difference test
of the two models showed that they were not redundant (χ
difference = 44.39, df = 6). Because the
four-factor solution provided better data fit indices and is more theoretically sound, we concluded it
was the best representation of the constructs underlying dimensionality.
Both two-factor and three-factor models also were tested and compared with the four-factor target
model. The two-factor model with the social astuteness and networking ability items comprising one
factor and the interpersonal influence and apparent sincerity items making up a second factor showed
lower-than-recommended fit statistics and, along with the significant chi-square difference test (χ
ference = 37.78, df = 5), indicates that the four-factor model is superior to the two-factor model.
The same three-factor model as tested in Study 1 was examined here as well, where networking
ability, social astuteness, and interpersonal influence reflected the three dimensions, with the three
items from the apparent sincerity dimension spread similarly across these three factors. The fit statis-
tics for the three-factor model and the significant chi-square difference test (χ
difference = 46.09, df =
3) indicates that the four-factor model also provides a better fit than the three-factor model.
Each of the four political skill dimensions demonstrated adequate reliability estimates. Specifi
cally, the social astuteness, interpersonal influence, networking ability, and apparent sincerity dimen
sions achieved reliability coefficients of .79, .77, .83, and .73, respectively. All items demonstrated
acceptable standardized factor loadings with their hypothesized underlying construct, and the four
dimensions exhibited correlations with one another, ranging from .28 to .69, as noted above the
diagonal in Table 2.
Sample 2: Sample and Procedure
A total of 148 full-time workers in law firms in a large southeastern city were contacted to complete
an online survey that provided data to test political skills convergent and discriminant validity. Ninety-
three (93) employees in 49 different legal entities voluntarily completed surveys online by accessing
the researchers Web site, reflecting a 63% response rate. Respondents occupied a wide variety of posi
tions, including legal administrators, paralegals, legal assistants, secretaries, attorneys, and office
managers. The average age of the respondents in this sample was 41.0 years (σ = 10.05), 84% of the
sample was female, 58% had at least a 4-year college degree, and the majority was Caucasian (94%).
The average tenure was 7.2 years (σ = 6.46) with the organization, and 44% supervised at least one
other employee as part of their job responsibilities.
Ferris et al. / Political Skill Inventory 141
Sample 2: Measures
Political skill. The 18-item PSI developed in Study 1 was used to measure political skill and its
dimensions. For this study, the 7-point response format was used. The coefficient alpha internal con
sistency reliability estimate for the overall scale was .89, and the reliabilities for the political skill
dimensions were as follows: networking ability (.87), interpersonal influence (.87), social astuteness
(.80), and apparent sincerity (.58).
Self-monitoring. Self-monitoring was measured with the 18-item instrument presented in Snyder
(1987). The coefficient alpha reliability for this scale was .76.
Conscientiousness. Conscientiousness was examined using items from McCrae and Costa (1987).
The coefficient alpha reliability estimate for the 14-item conscientiousness scale was .82.
Influence tactics. Kipnis et al.s (1980) measure was used to assess the frequency with which indi
viduals employed upward influence tactics. Each of the influence tactics scales had four items and
included upward appeal (α = .82), coalition (α = .80), and assertiveness (α = .77).
Trait anxiety. Trait anxiety was measured with the 20-item Spielberger et al. (1983) State-Trait
Anxiety Inventory (Form Y-2). The coefficient alpha reliability for this scale was .87.
Political savvy. A six-item scale developed by Chao et al. (1994) was used, which assessed under-
standing or savvy about politics in the organization. The coefficient alpha reliability estimate for this
scale was .85.
Sample 2: Convergent and Discriminant Validity Results
Table 5 shows the descriptive statistics for the variables used in the current study and the correla-
tions of political skill (and its dimensions) with each of the other scales. Among all the variables, only
the apparent sincerity political skill dimension (α = .58) had an inferior internal consistency reliability
We expected political skill to be positively correlated with self-monitoring and with conscientious
ness. The overall PSI was found to positively correlate with self-monitoring (r = .33, p < .01). Interest
ingly, although positive, the correlation of the PSI with conscientiousness was not significant in this
study. Furthermore, political skill should be positively correlated with the Chao et al. (1994) measure,
which assesses political savvy. As shown in Table 5, the overall PSI was found to positively correlate
with political savvy (r = .47, p < .001), thus providing partial support for Hypothesis 1.
Social astuteness demonstrated the highest correlation with self-monitoring (r = .32, p < .001), but
this correlation was not significantly different from the network buildingself-monitoring relationship
(i.e., t = .22, n.s.). Social astuteness also was hypothesized to demonstrate the highest correlation with
conscientiousness of any of the PSI dimensions, and tests for the difference of correlations verify this,
thus providing support for Hypothesis 2 (i.e., t = 2.03, p < .05 for interpersonal influence; t = 3.38,
p < .001 for network building; and t = 1.55, p < .10 for apparent sincerity).
Hypothesis 2 also proposed that social astuteness would demonstrate the strongest relationship
with political savvy of any of the four political skill dimensions, and the results appear to support this
(r = .60, p < .001). Furthermore, the results of the correlation difference tests demonstrate that the
social astutenesspolitical savvy correlation is significantly larger than the correlation of political savvy
with any other PSI dimension (i.e., t = 7.83, p < .001 for interpersonal influence; t = 2.76, p < .01 for
network building; and t = 5.76, p < .001 for apparent sincerity).
142 Journal of Management / February 2005
Descriptive Statistics and Correlations of Political Skill With Other Measures (Study 2, Sample 2) (N = 93)
Political Social Interpersonal Networking Apparent
Measures MSDα Skill Total Astuteness Influence Ability Sincerity
Self-monitoring 1.42 0.20 .76 .33** .32** .21* .30** .13
Conscientiousness 5.75 0.91 .82 .17 .31** .15 .00 .15
Upward appeal 1.42 0.58 .82 .17 .10 .00 .26* .03
Coalition 1.92 0.75 .80 .28** .23* .12 .30** .10
Assertiveness 1.52 0.70 .77 .16 .16 –.04 .22* .01
Trait anxiety 1.69 0.35 .87 –.27** –.25** –.42** –.11 –.19
Political savvy 5.49 1.09 .85 .47** .60*** .14 .38** .11
M 5.52 5.29 5.87 5.00 6.49
SD 0.71 0.90 0.82 1.13 0.46
α .89 .80 .82 .87 .58
†p < .10
*p < .05
**p < .01
***p < .001
Political skill should be related to various influence tactics, including upward appeal and coalition
tactics, but not to assertiveness, as was found in Study 1. Results provided evidence that political skill
is not redundant with influence tactics. Specifically, the PSI positively correlated with the coalition
tactic (r = .28, p < .01), and it did not correlate with assertiveness (r = .16, n.s.), both findings support
ive of Hypothesis 3. However, the PSI exhibited a positive, but not a significant, correlation with the
upward appeal tactic, which fails to support part of Hypothesis 3.
The networking ability dimension was positively related to the upward appeal influence tactic (r =
.26, p < .05), the coalition tactic (r = .30, p < .01), and the assertiveness tactic (r = .22, p < .05). How
ever, examination of the correlation difference tests indicate mixed results and therefore only partial
support for Hypothesis 4. The network buildingupward appeal correlation was significantly greater
than for any of the other PSI dimensions (i.e., t = 1.68, p < .05 for social astuteness; t = 2.56, p < .01 for
interpersonal influence; and t = 1.88, p < .05 for apparent sincerity). For the coalition influence tactic,
the network building correlation was not significantly different from the social astuteness correlation
with the coalition tactic (i.e., t = .74, n.s.), and the networking-assertiveness correlation was not signif
icantly different from the social astutenessassertiveness relationship (i.e., t = .62, n.s.).
In addition, political skill should be negatively related to trait anxiety such that politically skilled
individuals are apt to experience less anxiety or tension. Results show that the PSI indeed was
inversely related to trait anxiety (r = .27, p < .01), as found in Study 1, and supportive of Hypothesis 5.
Also replicating results from Study 1, interpersonal influence showed the strongest relationship with
trait anxiety (r = .42, p < .01), and correlations of trait anxiety with all other PSI dimensions were
found to be significantly smaller in magnitude, thus supporting Hypothesis 6 (i.e., t = 2.26, p < .01 for
social astuteness; t = 3.22, p < .001 for network building; and t = 2.59, p < .01 for apparent sincerity).
Sample 3: Participants and Procedure
A sample of 184 undergraduate students in two required undergraduate business classes at a
midsize southeastern university completed surveys during class time. All students present in class on
that day participated in the research. The average age of the respondents in this sample was 21.8 (σ =
2.74), 49% of the sample were female, and 86% were Caucasian.
Sample 3: Measures
Political skill. Political skill (M = 5.68, σ = .61, α = .87), again, was measured with the 18-item PSI.
The descriptive statistics for the four dimensions of political skill are as follows: social astuteness, M =
5.77, σ = .69, α = .71; interpersonal influence, M = 5.51, σ = .77, α = .73; networking ability, M = 5.82,
σ = .68, α = .76; and apparent sincerity, M = 5.52, σ = .95, α = .66.
General mental ability. GMA was measured using the Wonderlic Personnel Test Form A
(Wonderlic Personnel Test, 1992). The mean score was 25.76 (σ = 5.2).
Emotional intelligence. Emotional intelligence was measured using the Emotional Competence
InventoryUniversity Edition (ECI-U; Goleman & Boyatzis, 2001). Mayer, Salovey, and Caruso
(2000) described this scale as a mixed-model scale in that it goes beyond the emotions and emotion-
thought interactions and incorporates other characteristics such as consciousness states, abilities, and
motivations (M = 3.84, σ = .48, α = .96).
144 Journal of Management / February 2005
Sample 3: Results
With all the attention the emotional intelligence construct has received in recent years, it was impor
tant to demonstrate that it was not highly correlated with political skill. We would expect a significant
correlation between political skill and emotional intelligence, but we would not expect the magnitude
of the correlation to be so high as to indicate construct redundancy. The correlation between these two
constructs was .53, providing support for Hypothesis 1. Furthermore, the correlations of the four
dimensions of political skill with emotional intelligence ranged from .38 to .43.
Based on these results, two observations deserve mention. First, we would conclude that this level
of relationship is only moderate in magnitude. Second, in light of the extensiveness regarding how
Goleman and his colleagues have conceptualized the content domain of emotional intelligence (i.e., as
reflected in the measure used in this study), it could be construed as surprising that the correlation with
political skill was not even higher. Some have suggested that Goleman views emotional intelligence as
everything except GMA (e.g., Hedlund & Sternberg, 2000), suggesting a measure that would overlap
substantially with a number of social effectiveness constructs.
It is critical to demonstrate that political skill is uncorrelated with, and therefore not simply sub
sumed by, GMA. The correlations of the PSI total score, and the four dimensions, with GMA all were
found to be zero. This supports the emerging theory on political skill, as well as the evidence provided
by Ferris et al. (1999) on the initial six-item measure, which was found to reflect a zero correlation
with GMA (i.e., also using the Wonderlic Personnel Test to assess intelligence). Also, these results
provide support for Hypothesis 7.
Study 3: Criterion-Related Validity Evidence
The establishment of sound evidence of construct validity is an essential first step in the develop-
ment of any new measure, before such measures can be used in substantive research (e.g., Schwab,
1980). However, criterion-related validity ultimately is critical to the determination of any new con-
struct’s role in the predictability of important organizational phenomena.
To examine the criterion-related validity of political skill and provide tests of Hypotheses 8 and 9,
samples were gathered from two organizations, involving different occupational groups, to assess the
extent to which political skill predicted job performance and effectiveness ratings. In both samples,
employees completed questionnaire measures, including political skill, and job performance or effec
tiveness ratings were filled out by other sources (i.e., either the subordinates or the supervisors of those
Sample 1: Sample and Procedures
This sample was collected from school administrators in a public school district located in the mid
western United States. The survey was distributed to 35 administrators through interdepartmental
mail, 26 surveys were returned for a 74% response rate, and participation was voluntary. The respon
dents were all Caucasian, and predominantly male (62%), with most holding a masters or higher
degree (73.1%). The average respondent was 48 years old with an average of 7.1 years of
Sample 1: Measures
Political skill was measured using the 18-item PSI as discussed earlier. Respondents were asked to
rate their agreement with the questions on a scale of 1 (strongly disagree)to7(strongly agree). Three
Ferris et al. / Political Skill Inventory 145
items were used to measure leader effectiveness, gathered from responses by the subordinates or direct
reports of each school administrator: (a) Our manager is effective in representing the work unit to
upper management, (b) Our manager is effective in meeting the job-related needs of work unit mem
bers, and (c) Our manager is effective in meeting the needs of the organization. The coefficient alpha
for this three-item scale was .85 (M = 5.60, σ = .73).
Sample 1: Results
Regression analyses were used to assess the ability of political skill to explain variance in ratings of
leader effectiveness. In the first regression analysis, the PSI total score was found to account for a sig
nificant portion of the variance in leader effectiveness ratings (R
= .16, F(1, 24) = 4.64, p < .05), sup
porting Hypothesis 8. In the second regression analysis, the four PSI dimensions (i.e., social astute
ness, interpersonal influence, networking ability, and apparent sincerity) were used to predict ratings
of leader effectiveness. As proposed in Hypothesis 9, social astuteness was expected to demonstrate
the strongest relationship with supervisor ratings of employee job performance. Indeed, social astute
ness was found to be the only significant predictor of effectiveness ratings (β = .11, p < .05), thus pro
viding support for Hypothesis 9. However, the regression equation was not significant, F(4, 21) = 1.88,
n.s., likely due to the small sample size.
Although these results are suggestive of political skills criterion-related validity, they are limited by
the size of the sample and the lack of control variables. Therefore, a more refined test of the criterion-
related validity of political skill was conducted, which included a larger sample and additional vari-
ables to serve as controls, in an effort to replicate the results reported here.
Sample 2: Sample and Procedures
Surveys were mailed to 474 branch managers of a national financial services firm. An endorsement
letter from the vice president of Human Resources accompanied the survey. The survey packet also
contained an approach letter from the research team, explaining the purpose of the survey and ensuring
that their participation was both voluntary and confidential. Of the 474 surveys distributed, 148 total
responses were obtained (i.e., 31.2% response rate). The respondents were predominantly female
(80.7%) and Caucasian (64.8%), with representative African American (22.8%) and Hispanic (7.8%)
respondents. The average respondent was 37 years old (σ = 9.83) and had an average of 2.53 (σ = 1.81)
years of organizational tenure and 1.41 years (σ = 1.54) of tenure with their current supervisor.
Sample 2: Measures
The 18-item PSI was used to measure political skill. Respondents were asked to rate their agree
ment with the questions on a scale of 1 (strongly disagree)to5(strongly agree). The scale demon
strated acceptable reliability (α = .86, M = 3.89, σ = .38). To measure employee performance, the most
recent annual internal performance ratings were obtained from the Human Resources department of
the firm. Employees were rated by their immediate supervisor in 28 categories ranging from budgeted
revenue growth to interpersonal relationships. The rating form used a 5-point behavioral anchored
response format. Performance ratings were calculated by summing the scores in each category (M =
81.39, σ = 14.21).
Of the 148 responses obtained from the original survey, performance scores were available for 108
of the employees, reducing the usable response rate to 22.9%. To isolate the effects of political skill on
performance ratings, the current study controlled for the effects of age, tenure with supervisor, organi
146 Journal of Management / February 2005
zation tenure, self-monitoring (α = .78, M = 3.61, σ = .34), and the influence tactics of exchange (α =
.54, M = 1.79, σ = .70), coalition building (α = .71, M= 2.30, σ = .71), upward appeals (α = .70, M =
1.79, σ = .84), and assertiveness (α = .66, M= 1.67, σ = .69).
Sample 2: Results
On the first step of the hierarchical regression analysis, the control variables of age, supervisor ten
ure, organization tenure, self-monitoring, exchange tactics, upward appeal, and coalition tactics
accounted for a significant proportion of variability in performance, R
= .18, F(7, 100) = 3.08, p < .01.
Within this block of variables, only organization tenure demonstrated a significant relationship with
performance ratings (β = .31, p < .01). The addition of the PSI total score in the second block of vari
ables contributed significant incremental variance beyond the control variables, ∆R
= .04, F(1, 99) =
5.17, p < .05. As predicted, employee political skill was positively related to supervisor-rated job per
formance (β = .22, p < .05), which provides additional support for Hypothesis 8.
To assess how each of the dimensions of the PSI contributed as predictors, a hierarchical regression
analysis was conducted entering the four political skill dimensions, after the control variables, in the
prediction of job performance ratings. Again, we would expect social astuteness to be the strongest
predictor of job performance ratings. Entering the four PSI dimensions on the next step resulted in a
7% increment in the proportion of criterion variance explained, F(4, 96) = 2.19, p < .10, with only the
social astuteness dimension reaching significance (β = .26, p < .05). These results replicate the crite-
rion-related validity of social astuteness in the prediction of job performance found in the previous
study, support Hypothesis 9, and are consistent with the derivation and discussion of this dimension of
political skill presented.
The 18-item PSI was developed in this research and found to reflect adequate psychometric proper-
ties as indicated by the confirming evidence across three studies and seven different samples.
Evidence for factor structure. An underlying four-factor dimensionality of the PSI was established
in Study 1 and confirmed in Study 2, with acceptable fit statistics. The correlations among the political
skill dimensions were modest in magnitude, as expected, and recent research has supported such
Evidence for convergent and discriminant validity. Part of Study 1, and two of the samples in Study
2, focused on convergent and discriminant validity and supported three fundamental conceptual issues
regarding political skill. That is, first, that political skill is positively correlated with personality and
other interpersonally oriented constructs like self-monitoring, political savvy, and emotional intelli
gence, but not so highly as to indicate construct redundancy. Second, that political skill is significantly
and negatively related to trait anxiety, thus supporting arguments by Perrewé et al. (2000). Third, that
political skill is not correlated with GMA.
The apparent sincerity dimension of political skill did not distinguish itself as providing differential
prediction. The problem with this PSI dimension might be at least partially due to measurement
method. That is, the extent to which one is successful at appearing sincere is a function of others per
ceptions. Yet, the self-report measurement used in this research for political skill asked individuals to
report how much they intended to be sincere, which might have little to do with what others think of
their sincerity. This underscores the importance of using other sources for the measurement of ones
Ferris et al. / Political Skill Inventory 147
political skill rather than simply relying exclusively on self-reports. Furthermore, the reliability esti
mate of .58 for apparent sincerity in Sample 2 of Study 2 is lower than considered acceptable
(Nunnally, 1978) and places a lower bound on the possible validity coefficients that could be observed.
Evidence for criterion-related validity. Study 3 reported evidence of criterion-related validity and
demonstrated that political skill significantly predicted job performance and effectiveness ratings in
two samples made up of distinct occupational groups. The PSI was a significant predictor of subordi
nate evaluation of leader effectiveness for a group of public school administrators, even though the
sample was very small (i.e., N = 26), and social astuteness was the strongest PSI dimension predicting
effectiveness. These results were replicated in the second sample with supervisor ratings of job
performance as the criterion.
Limitations and Directions for Future Research
As with any empirical study, there are limitations regarding this research that should be mentioned.
Because of the largely single-source, self-report data collection, it is possible that common method
variance could explain some of the observed relationships of political skill with other constructs.
Therefore, in an attempt to empirically assess the potential problematic nature of common method
variance in this research, Harman one-factor tests were conducted in both Study 1 and Study 2 (i.e.,
including the four political skill dimensions, self-monitoring, conscientiousness, the three influence
tactics, and trait anxiety).
The basic notion is that if method variance is a serious problem, either one general factor will
account for most of the variance or a single factor will be found from the factor analysis (e.g.,
Podsakoff & Organ, 1986). In both studies, a single factor did not emerge from the factor analysis, nor
did a general factor account for the majority of variance (i.e., in Study 1, Factors 1 and 2 accounted for
28% and 27% of the variance, respectively; in Study 2, Factor 1 and Factor 2 explained 28% and 21%
of the variance, respectively). Therefore, common method variance does not appear to be a serious
problem in this research.
Further evidence that common method variance is not problematic in this research comes from an
examination of the correlations among the political skill dimensions (see Table 2) and the correlations
of political skill (i.e., total score and the four dimensions) with other constructs (see Tables 4 and 5).
Common method variance tends to produce a general spurious inflation of all the correlations among
variables measured with single-source data collection, resulting in an overestimation of the true rela
tionships. Clearly, this is not the case in examination of Tables 2, 4, and 5, which reflect reasonable and
expected variability concerning the magnitude of the relationships.
The two dimensions of political skill that appear to support the clearest differential prediction are
social astuteness and networking ability. First, social astuteness was proposed to be the only political
skill dimension to be predictive of job performance. Also, networking ability reflected a reasonably
clean differential relationship with the upward appeal, coalition, and assertiveness influence tactics in
Study 1, but with mixed findings in Study 2.
We envision two main roles for political skill in predicting job performance and other work out
comes. First, the PSI should be a main effect predictor, similar to results reported by Higgins (2000)
and Ahearn et al. (2004) using the earlier six-item measure. Indeed, an important area for future work
concerns the nature of leader political skill and how leader political skill operates to inspire trust and
confidence in followers, as suggested theoretically.
Second, we see political skill as a potentially important moderator that should facilitate the effec
tiveness of influence tactics on performance. Indeed, recent meta-analyses of the influence
tacticswork outcomes relationships have reported the existence of moderators (Gordon, 1996; Hig
gins, Judge, & Ferris, 2003). Furthermore, we would expect to see political skill act as a moderator of
148 Journal of Management / February 2005
the stress-strain relationship, serving as an antidote of sorts, as argued by Perrewé et al. (2000) and
reported by Perrewé et al. (2004), using the six-item Ferris et al. (1999) scale.
Ferris, Perrewé, et al. (2002) discussed the proliferation of social constructs (e.g., social skill, polit
ical skill, social intelligence, etc.) and the need to precisely delineate their individual uniqueness. They
argued that many of these constructs share in common a cognitive understanding or perceptiveness
component in addition to a behavioral action component used to act on the former knowledge and,
therefore, are all reflective of a higher-order construct we might refer to as social effectiveness. The
concept of charisma also would be part of this category, and it would be interesting to examine the
extent to which charisma is simply captured in political skill. Clearly, there is a need to examine the
relationships among some of these social effectiveness constructs.
Finally, additional work in this area also should consider collecting political skill data from sources
other than the job incumbent, as noted above in discussing some of the problems with the apparent sin
cerity dimension. Self-report measurement of interpersonally oriented constructs, like political skill,
can be useful and enlightening (e.g., Riggio & Riggio, 2001). However, if we were able to demonstrate
some consistent level of employee political skill agreement across sources (e.g., superiors, peers, and
subordinates), we would have greater confidence in the measures ability to tap meaningful elements of
Political Skill Item Pool
Instructions: Using the following 7-point scale, please place the number on the blank before each item that best
Instructions: describes how much you agree with each statement about yourself.
1 = strongly disagree
2 = disagree
3 = slightly disagree
4 = neutral
5 = slightly agree
6 = agree
7 = strongly agree
1. _____ I find it easy to envision myself in the position of others.
2. _____ In social situations, it is clear to me just what to say and do.
3. _____ I spend a lot of time and effort at work networking with others. (NA)*
4. _____ I am good at getting others to work well together.
5. _____ I am able to make most people feel comfortable and at ease around me. (II)
6. _____ I am good at making myself visible with influential people in my organization.
7. _____ I am able to adjust my behavior and become the type of person dictated by any situation.
8. _____ I am able to communicate easily and effectively with others. (II)*
9. _____ It is easy for me to develop good rapport with most people. (II)
10. _____ I am good at reading social situations, and determining the most appropriate behavior to demonstrate
the proper impression.
11. _____ I am very conscious of how I am perceived by others.
12. _____ I have always prided myself in having good savvy, street smarts, or political skill at work.
13. _____ I understand people very well. (SA)
14. _____ I am the one who can get people to work well together.
15. _____ I try to make people feel important by what I say and do.
16. _____ I am good at building relationships with influential people at work. (NA)*
17. _____ I am good at getting others to respond positively to me.
18. _____ I usually try to find common ground with others.
19. _____ I think a lot about how, as well as what, I say when presenting an idea to others.
20. _____ I size up situations before deciding how to present an idea to others.
Ferris et al. / Political Skill Inventory 149
21. _____ I am particularly good at sensing the motivations and hidden agendas of others. (SA)*
22. _____ I am good at reading others’ body language.
23. _____ When communicating with others, I try to be genuine in what I say and do. (AS)*
24. _____ I have developed a large network of colleagues and associates at work whom I can call on for support
when I really need to get things done. (NA)
25. _____ At work, I know a lot of important people and am well connected. (NA)*
26. _____ I spend a lot of time at work developing connections with others. (NA)*
27. _____ I try to get others to talk about themselves.
28. _____ I listen carefully and attentively when people talk to me.
29. _____ I am good at getting people to like me. (II)*
30. _____ It is important that people believe I am sincere in what I say and do. (AS)*
31. _____ I try to show a genuine interest in other people. (AS)*
32. _____ I am good at using my connections and network to make things happen at work. (NA)*
33. _____ I try to see others’ points of view.
34. _____ I try to find solutions to problems that incorporate others’ views and opinions.
35. _____ I am good at coordinating the efforts and talents of team members to bring about effective team
36. _____ I am conscious of getting myself in the best position to take advantage of opportunities.
37. _____ I have good intuition or savvy about how to present myself to others. (SA)*
38. _____ I always seem to instinctively know the right things to say or do to influence others. (SA)*
39. _____ I pay close attention to people’s facial expressions. (SA)*
40. _____ Sometimes I feel like an actor because I have to play different roles with different people.
Note: The asterisked items indicate retained items in the final 18-item scale. Items marked with † indicate the original six items
developed by Ferris et al. (1999). NA = networking ability; II = interpersonal influence; SA = social astuteness; AS = apparent
Ahearn, K. K., Ferris, G. R., Hochwarter, W. A., Douglas, C., & Ammeter, A. P. 2004. Leader political skill and team perfor-
mance. Journal of Management. 30: 309-327.
Anderson, J. C., & Gerbing, D. W. 1988. Structural equation modeling in practice: A review and recommended two-step
approach. Psychological Bulletin, 103: 411-423.
Bacharach, S. B., & Lawler, E. J. 1998. Political alignments in organizations: Contextualization, mobilization, and coordination.
In R. M. Kramer & M. A. Neale (Eds.), Power and influence in organizations: 67-88. Thousand Oaks, CA: Sage.
Buck, R. 1991. Temperament, social skills, and the communication of emotion: A developmental-interactionist view. In D. G.
Gilbert & J. J. Connolly (Eds.), Personality, social skills, and psychopathology: An individual difference approach: 85-105.
New York: Plenum.
Chao, G. T., O’Leary-Kelly, A. M., Wolf, S., Klein, H. J., & Gardner, P. D. 1994. Organizational socialization: Its content and
consequences. Journal of Applied Psychology, 79: 730-743.
Cohen, J., & Cohen, P. 1983. Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale,
NJ: Lawrence Erlbaum.
Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. 1999. Evaluating the use of exploratory factor analysis in psy
chological research. Psychological Methods, 4: 272-299.
Ferris, G. R., Anthony, W. P., Kolodinsky, R. W., Gilmore, D. C., & Harvey, M. G. 2002. Development of political skill. In
C. Wankel & R. DeFillippi (Eds.), Research in management education and development, Volume 1: Rethinking management
education for the 21st century: 3-25. Greenwich, CT: Information Age Publishing.
Ferris, G. R., Berkson, H. M., Kaplan, D. M., Gilmore, D. C., Buckley, M. R., Hochwarter, W. A., et al. 1999. Development and
initial validation of the political skill inventory. Paper presented at the 59th annual national meeting of the Academy of Man
Ferris, G. R., Hochwarter, W. A., Douglas, C., Blass, F. R., Kolodinsky, R. W., & Treadway, D. C. 2002. Social influence pro
cesses in organizations and human resources systems. In G. R. Ferris & J. J. Martocchio (Eds.), Research in personnel and
human resources management, Vol. 21: 65-127. Oxford, UK: JAI/Elsevier Science.
Ferris, G. R., Perrewé, P. L., & Douglas, C. 2002. Social effectiveness in organizations: Construct validity and research direc
tions. Journal of Leadership and Organizational Studies, 9: 49-63.
150 Journal of Management / February 2005
Gefen, D., Straub, D., & Boudreau, M. C. 2000. Structural equation modeling and regression: Guidelines for research practice.
Communications of the Association for Information Systems, 4: 1-78.
Goleman, D. 1995. Emotional intelligence. New York: Bantam Books.
Goleman, D. 1998. Working with emotional intelligence. New York: Bantam Books.
Goleman, D., & Boyatzis, R. 2001. Emotional competence inventory–University edition. Boston: Hay Group, Inc.
Gordon, R. A. 1996. Impact of ingratiation on judgments and evaluations: A meta-analytic investigation. Journal of Personality
and Social Psychology, 71: 54-70.
Gottfredson, L. S. 1997. Why g matters: The complexity of everyday life. Intelligence, 24: 79-132.
Hair, J., Anderson, R. O., & Tatham, R. 1987. Multidimensional data analysis. New York: Macmillan.
Hatcher, L. 1994. A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. Cary,
NC: SAS Institute.
Hedlund, J., & Sternberg, R. J. 2000. Too many intelligences? Integrating social, emotional, and practical intelligence. In R. Bar-
On & J. Parker (Eds.), The handbook of emotional intelligence: 136-167. San Francisco: Jossey-Bass.
Higgins, C. A. 2000. The effect of applicant influence tactics on recruiter perceptions of fit. Unpublished doctoral dissertation,
Department of Management and Organizations, University of Iowa.
Higgins, C. A., Judge, T. A., & Ferris, G. R. 2003. Influence tactics and work outcomes: A meta-analysis. Journal of Organiza
tional Behavior, 24: 89-106
Hinkin, T. R. 1995. A review of scale development practices in the study of organizations. Journal of Management, 21: 967-988.
Hu, L., & Bentler, P. M. 1999. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new
alternatives. Structural Equation Modeling, 6: 1-55.
Jones, E. E. 1990. Interpersonal perception. New York: Freeman.
Jöreskog, K., & Sörbom, D. 1993. LISREL 8 users guide. Uppsala, Sweden: University of Uppsala.
Kipnis, D., Schmidt, S. M., & Wilkinson, I. 1980. Intraorganizational influence tactics: Explorations in getting ones way. Jour
nal of Applied Psychology, 65: 440-452.
La Du, T. J., & Tanaka, J. S. 1989. Influence of sample size, estimation method, and model specification on goodness of fit
assessment in structural equation models. Journal of Applied Psychology, 74: 625-635.
Lewis-Beck, M. S. 1980. Applied regression: An introduction. Beverly Hills, CA: Sage.
Luthans, F., Hodgetts, R. M., & Rosenkrantz, S. A. 1988. Real managers. Cambridge, MA: Ballinger.
Mayer, J. D., Salovey, P., & Caruso, D. R. 2000. Models of emotional intelligence. In R. J. Sternberg (Ed.), The handbook of intel-
ligence: 396-420. New York: Cambridge University Press.
McCrae, R., & Costa, P., Jr. 1987. Validation of the five-factor model of personality across instruments and observers. Journal of
Personality and Social Psychology, 52: 81-90.
Medsker, G. J., Williams, L. J., & Holohan, P. J. 1994. A review of current practices for evaluating causal models in organiza-
tional behavior and human resources management research. Journal of Management, 20: 439-464.
Mintzberg, H. 1983. Power in and around organizations. Englewood Cliffs, NJ: Prentice Hall.
Mintzberg, H. 1985. The organization as a political arena. Journal of Management Studies, 22: 133-154.
Mulaik, S. A., James, L. R., Van Alstine, J., Bennet, N., Lind, S., & Stillwell, C. D. 1989. An evaluation of goodness-of-fit indi
ces for structural equation models. Psychological Bulletin, 105: 430-445.
Murtha, T. C., Kanfer, R., & Ackerman, P. L. 1996. Toward an interactionist taxonomy of personality and situations: An integra
tive situational-dispositional representation of personality traits. Journal of Personality and Social Psychology, 71: 193-207.
Nunnally, J. C. 1978. Psychometric theory (2nd ed.). New York: McGraw-Hill.
Perrewé, P. L., Ferris, G. R., Frink, D. D., & Anthony, W. P. 2000. Political skill: An antidote for workplace stressors. Academy of
Management Executive, 14: 115-123.
Perrewé, P. L., Zellars, K. L., Ferris, G. R., Rossi, A. M., Kacmar, C. J., & Ralston, D. A. 2004. Neutralizing job stressors: Politi
cal skill as an antidote to the dysfunctional consequences of role conflict stressors. Academy of Management Journal, 47:
Pfeffer, J. 1981. Power in organizations. Boston: Pitman.
Pfeffer, J. 1992. Managing with power: Politics and influence in organizations. Boston: Harvard Business School Press.
Podsakoff, P. M., & Organ, D. W. 1986. Self-reports in organizational research: Problems and prospects. Journal of Manage
ment, 12: 531-544.
Riggio, R. E., & Riggio, H. R. 2001. Self-report measurement of interpersonal sensitivity. In J. A. Hall & F. J. Bernieri (Eds.),
Interpersonal sensitivity: Theory and measurement: 127-142. Mahwah, NJ: Lawrence Erlbaum.
Schwab, D. 1980. Construct validity in organizational behavior. In B. M. Staw & L. L. Cummings (Eds.), Research in organiza
tional behavior, Vol. 2: 3-43. Greenwich, CT: JAI.
Snyder, M. 1987. Public appearances, private realities: The psychology of self-monitoring. New York: Freeman.
Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A. 1983. Manual for the State-Trait Anxiety Inventory
STAI (Form Y). Palo Alto, CA: Consulting Psychologists Press.
Strahan, R., & Gerbasi, K. C. 1972. Short, homogenous version of Crowne-Marlowe Social Desirability Scale. Journal of Clini
cal Psychology, 28: 191-193.
Ferris et al. / Political Skill Inventory 151
Tucker, L. R., & Lewis, C. 1973. A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38: 1-10.
Wheaton, B., Muthen, B., Alwin, D., & Summers, G. 1977. Assessing reliability and stability in panel models. In D. Heise (Ed.),
Sociological methodology: 84-136. San Francisco: Jossey-Bass.
Wonderlic Personnel Test, Inc. 1992. Wonderlic personnel test and scholastic level exam: Users manual. Libertyville, IL:
Wonderlic Personnel Test.
Gerald R. Ferris is the Francis Eppes Professor of Management and a professor of psychology at Florida State University. He
received a Ph.D. in business administration from the University of Illinois at Urbana-Champaign.
Darren C. Treadway is an assistant professor of management at the University of Mississippi. He received a Ph.D. in manage
ment from Florida State University.
Robert W. Kolodinsky is an assistant professor of management at James Madison University. He received a Ph.D. in manage
ment from Florida State University.
Wayne A. Hochwarter is an associate professor of management at Florida State University. He received a Ph.D. in management
from Florida State Univerrsity.
Charles J. Kacmar is an associate professor of management information systems at the University of Alabama. He received a
Ph.D. in computer science from Texas A&M University.
Ceasar Douglas is an assistant professor of management at Florida State University. He received a Ph.D. in business administra-
tion from the University of Mississippi.
Dwight D. Frink is an associate professor of management and the PMB and William King Self Chair at the University of Missis-
sippi. He received a Ph.D. in human resources management from the Institute of Labor and Industrial Relations at the University
of Illinois at Urbana-Champaign.
152 Journal of Management / February 2005