Influence Tactics and Work Outcomes: A Meta-Analysis
Author(s): Chad A. Higgins, Timothy A. Judge, Gerald R. Ferris
Source: Journal of Organizational Behavior, Vol. 24, No. 1 (Feb., 2003), pp. 89-106
Published by: John Wiley & Sons
Stable URL: http://www.jstor.org/stable/4093798
Accessed: 21/01/2010 03:32
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=jwiley.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
John Wiley & Sons is collaborating with JSTOR to digitize, preserve and extend access to Journal of
Organizational Behavior.
http://www.jstor.org
Journal of OrganizationalBehavior
J. Organiz.Behav. 24, 89-106 (2003)
Publishedonline in Wiley InterScience(www.interscience.wiley.com).DOI: 10.1002/job.181
Influence tactics and work outcomes:
a meta-analysis
CHAD A. HIGGINS'*, TIMOTHY A. JUDGE2 AND GERALD R. FERRIS3
'Departmentof Managementand Organization,Universityof Washington,U.S.A.
2Departmentof Management,Universityof Florida, U.S.A.
3Departmentof Management,Florida State University,U.S.A.
Summary
Recentresearchoninfluencetacticshasfocusedon thebenefitsthataccrueas a resultof using
suchtactics.Thecurrentstudyutilizesmeta-analytic
techniquesto estimatethetruepopulationcorrelations
betweenvariousinfluencetacticsandwork-related
outcomes.Resultsindicate thatingratiationand rationalityhave positiveeffects on workoutcomes.Additional
effectsin certainsituaanalysessuggestthattheseandotherinfluencetacticshavesignificant
tionsandon specificworkoutcomes.Copyright0 2003 JohnWiley& Sons,Ltd.
Introduction
Over 40 years ago, Goffman (1955) introducedto the behavioralsciences the notion thatpeople consciously managethe impressionsthey convey to othersin interpersonalinteractions.While manipulation was not a new concept at that time, it was indeed a novel argumentthat people act out roles in
efforts to establish identities they wish to convey, and which can result in personal gain. It was also
intriguingto suggest thatpeople alterthe image they choose to present,andthe strategyused to present
this image, based on the situationthey are in and the outcomes they hope to achieve. Goffman'sdramaturgicalperspectivestronglyinfluencedthe evolution of theory and researchin interpersonalbehavior, and subsequentstreamsof relatedwork on self-presentation,impressionmanagement,influence
tactics, and organizationalpolitics, have remainedactive areas of inquiry to date.
Somewhatlater, within the past two decades, interpersonalinfluence theory was applied to organizational settingsin effortsto betterunderstandthe behaviorof people at work.Worksettings were seen
as conducive to interpersonalinfluence because obtaining a job offer, receiving a favorable performance evaluation,or ensuring a high pay increase representrewardsat work which could stimulate
the self-interest motive and thus behaviors designed to manage favorable impressions. However, it
is importantto note that individualsdo not necessarily use the same influence strategyin every situation. Likewise, differentindividualsmay choose differentinfluence strategieswhen faced with similar
situations.Forexample,whereasone individualmay use self-promotionto obtainajob offer,that same
*
Correspondenceto: Chad A. Higgins, Departmentof Managementand Organization,School of Business, University of
Washington,Box 353200, Seattle, WA 98195. U.S.A. E-mail: chiggins@u.washington.edu
Copyright? 2003 JohnWiley& Sons,Ltd.
Received 23 April 2002
Revised 10 September2002
Accepted 16 September2002
90
C. A. HIGGINSETAL.
individualmay use ingratiationor rationalityin an attemptto obtaina promotionor pay raise. On the
otherhand,anotherindividual,when faced with the same situation,may use ingratiationto obtainajob
offer and assertivenessor self-promotionto win a pay raise.
A numberof contextualfactorsand individualdifferencesdeterminewhich influencetactics an individual chooses to use, underwhat circumstanceshe or she chooses to use them, and how effective the
tactic of choice will be. Such factorsinclude the relativepowerof the parties,the directionof the influence attempt,the objective of the influence attempt,and the political skill of the influencer(Falbe &
Yukl, 1992; Ferris,Perrewe,Anthony,& Gilmore, 2000). Anotherfactor affecting the success of an
influence attemptis the choice of tactic. Falbe and Yuklfound thatusing a single soft tactic (i.e., those
thatrely on personalpowerandpower sharing,such as ingratiationor consultation)was more effective
than a single hardtactic (i.e., those that rely on authorityand position power, such as self-promotion).
Furthermore,combining two soft tactics, or a soft tactic and rationality,was more effective than any
single tactic or a combinationof hardtactics. Thus, evidence suggests that differenttactics do have
differing degrees of effectiveness.
However,despite the fact researchershave investigatedthe effects of influencetactics on work outcomes for some time now, there has been little attentiondevoted to obtaininga comprehensiveassessment of the effects of influence tactics by cumulating results across studies. In addition, previous
attemptsto meta-analysethe effects of influencetactics have sufferedfrom several shortcomingsthat
limit the confidence one can have in the results of such studies.
Purpose of the present study
The purposeof the presentstudy is twofold. First,despite the increasedamountof researchexamining
the role of various influence tactics in organizations,the results of this researchreveal a lack of consistent findings.For example, whereas Orpen(1996) found a positive effect of ingratiationon promotions, Thackerand Wayne(1995) found a negative effect for the same relationship.Similarly,Stevens
and Kristof(1995) found a positive relationshipbetween self-promotionandperformanceassessments,
but Ferris,Judge, Rowland,and Fitzgibbons(1994) found self-promotionto have a negative effect on
performance.As a resultof these conflicting findings, it is difficultto determinethe true natureof the
relationshipbetween influencetactics and workoutcomes. In orderfor a literatureto advanceand have
an impacton the field of organizationalbehavior,it is importantthat we be able to show consistency in
researchfindings or, at the very least, attemptto explain discrepancies.Therefore,one purposeof the
currentstudy is to cumulatethe results of recent researchexamining the effect of influence tactics on
work outcomes, meta-analyticallyexamine these results,and try to provide a consistentexplanationof
these resultsthat will allow us to betterunderstandthe effects of influencetactic use in organizations.
The second purposeof the presentstudy is to overcomelimitationsof previousmeta-analyticstudies
of influence tactics. In particular,we seek to expand upon Gordon's(1996) recently publishedmetaanalysis of ingratiationstrategies.Although Gordon'sstudy providedthoroughcoverage of the ingratiation literature,there are several limitationsto this study that warrantfurtherstudy. First, there are
many influencetactics otherthaningratiationthatmay affect organizationaloutcomes. A review of the
influenceliteratureindicatedthatmost researchon influencetactics in organizationsfocused on one or
more of the following tactics: ingratiation, self-promotion, rationality, assertiveness, exchange,
upwardappeal,and/orcoalitions. In addition,Gordon'sinvestigationincludedself-promotionas a tactic of ingratiation.However, Godfrey,Jones, and Lord (1986) providedempirical evidence that selfpromotionand ingratiationare indeed distinct influence tactics and should be treatedas such. Therefore, it is theoretically and conceptually importantto distinguish self-promotionand ingratiationas
independentinfluence tactics.
Copyright( 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
INFLUENCEAND OUTCOMES
91
In addition,Gordon's(1996) meta-analysiswas based on a fixed effects method, which assumes no
between studies variance. However, based on the significant homogeneity statistics reported by
Gordon,this assumptiondid not hold true for the studies included in Gordon'smeta-analysis.Therefore, it is difficultto determinewhetheror not the effects reportedby Gordonare non-zero. Also, the
meta-analytictechniqueused by Gordon(1996) correctedeffect sizes only for the bias introducedby
the d-statistic'soverestimationof populationeffect sizes. This is a serious deficiency as measurement
erroroften accountsfor up to 50 per cent of the total varianceof a measure(Schmidt& Hunter,1996).
In the presentstudy,effect sizes were correctednot only for bias due to samplingerror,but also for bias
due to measurementerrorin both the independentand dependentvariables.
In light of the issues just raised, an updatedreview of the influence tactic literatureis necessary to
enhance our understandingof influence in organizations.The present study attemptsto addresssome
of the concernswith previousmeta-analyticreviews of influencetactics and contributeto the literature
by including a broaderset of influence tactics, investigating a comprehensiveframeworkof workoriented outcomes, and using appropriatemeta-analytictechniques.
Interpersonalinfluencetheory
Theory and researchon interpersonalinfluence has been developed and advancedlargely throughthe
contributionsof scholars such as Jones (Jones, 1964, 1990; Jones & Pittman, 1982), Tedeschi (1981),
and more recently,by Leary (1995; Leary & Kowalski, 1990). This body of work has articulatedthe
precise natureof interpersonalinfluence tactics, their antecedents,boundaryconditions, and consequences, and considerable empirical work has been generated in efforts to systematically examine
these theoreticalarguments.
Within the last two decades, interestemerged in applying interpersonalinfluence theory to organizational settings. The measurementof influence tactics in organizationalsettings began with the categorizationscheme proposedby Kipnis, Schmidt,and Wilkinson (1980), which identifiedeight tactics.
These tactics included assertiveness (using a forceful manner to get what one wants), ingratiation
(using behaviorsdesigned to increase the target'sliking of oneself or to make oneself appearfriendly
in order to get what one wants), rationality(using data and informationto make a logical argument
supportingone's request), sanctions (using punishmentor the threat of punishmentto gain compliance), exchange (makingan explicit offer to do somethingfor anotherin exchangefor theirdoing what
one wants), upwardappeals (relying on the chain of command, calling in superiorsto help get one's
way), blocking (attemptingto stop the target from carryingout some action by impeding their progress), and coalitions (mobilizing others to help in persuading the target individual). Kipnis and
Schmidt (1982) published scales for their principal influence tactics that have been used in several
organizationalstudies (e.g., Wayne, Liden, Graf, & Ferris, 1997). However, subsequentresearchon
the Kipniset al. (1980) tactics has omittedtwo of the tactics (sanctionsandblocking) due to conceptual
problems and infrequentuse (Kipnis & Schmidt, 1988; Yukl & Falbe, 1990).
In addition to the tactics put forth by Kipnis et al. (1980), Jones and Pittman (1982) proposed
anotherset of influence tactics believed to be relevantto interpersonalbehaviorin organizationalsettings. Jones and Pittman'scategorizationof self-presentationtactics was the first to distinguish the
tactic of self-promotion(attemptingto create an appearanceof competence or that you are capable
of completing a task) from ingratiation.Although Jones and Pittman'scategorizationincluded a total
of five self-presentationtactics, only self-promotionand ingratiationhave received substantialattention in the literature.
As widely studied as influence tactics are, there is little agreementas to the exact labels to be used
for each tactic. For example, Wayne and Ferris (1990) conducted a study of supervisor-focusedand
Copyright ) 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
92
C. A. HIGGINSETAL.
job-focused influencetactics. These tactics are essentially the same as ingratiationand self-promotion
respectively, but were identified by different labels in the Wayne and Ferris study. Therefore,in an
effort to provide consistency throughoutthe present study, we have adoptedthe Kipnis et al. (1980)
labels of assertiveness,coalitions, exchange, ingratiation,rationality,and upwardappeal,as well as the
self-promotionlabel proposedby Jones and Pittman(1982). Because they representa broad segment
of the influencetactic literatureand because of their widespreadacceptanceand use, these tactics will
be the focus of the present study.
Interpersonalinfluence in organizations
In 1988, Gardnerand Martinkoconducted a thoroughreview of the literatureon interpersonalinfluence. On the one hand, Gardnerand Martinko's(1988) review suggested that researchershad developed a clear understandingof the antecedentsto influence tactic use in organizations.On the other
hand, Gardnerand Martinkoalso noted that little researchhad examined the outcomes of, or behavioral responses to, influence strategies. Fortunately,researchersrecognized the importanceof this
area of enquiry and sought to addressthis issue.
Following the seminal work of Gardnerand Martinko(1988), much of the researchon interpersonal
influencein organizationsappliedinterpersonalinfluenceto humanresourcesmanagementsystems and
decisions (Ferris& Judge, 1991), with particularreferenceto selection andemploymentinterviewdecisions (e.g., Gilmore& Ferris, 1989; Gilmore,Stevens, Harrell-Cook,& Ferris, 1999), the performance
evaluationprocess and outcomes as engineeredby supervisors/raters
(e.g., Cleveland& Murphy,1992;
Villanova& Bernardin,1991) as well as by subordinates/ratees
(e.g., Ferris& Frink, 1997), the use of
as
goals (in goal setting) impressionmanagementtactics (e.g., Frink& Ferris,1998; Huber,Latham,&
Locke, 1989), and the career progression and success process, focusing on promotions and salary
increases (e.g., Cooper,Graham,& Dyke, 1993; Ferris,Fedor,& King, 1994; Judge & Bretz, 1994).
Factors affecting overall validity estimates
Because of the complex natureof interpersonalinteractionsand because of the even greatercomplexity introducedthroughthe use of influence tactics, a numberof variablescould potentially affect the
relationshipsbetween influence tactics and work outcomes. In the present study, the variablesmost
likely to affect such relationshipsinclude the specific work outcome of interest,the source of performance assessments, and the researchenvironment.Each of these will be discussed in turn.
Work outcomes
The specific work outcome of interestis likely to affect the relationshipbetween influencetactics and
work outcomes. Unlike performanceassessments, which are largely dependenton the behaviorof the
individual,measuresof extrinsicsuccess (e.g., salariesandpromotions)are often affectedby a number
of factorsthatare not directlyunderthe controlof the employee. For example, seniority,job level, and
genderaffect salaries(Gerhart& Rynes, 1991; Lawler, 1966), whereaspromotionsare at least partially
dependent upon the availability of open positions to be promotedto. Furthermore,because performance assessments typically occur more frequentlythan eitherpay raises or promotion,they are also
likely to occur closer in time to an individual'suse of influence tactics. As a result, there is a lesser
probabilitythatexternalfactors may interveneand weaken the effect the influencetactic(s) may have.
Therefore,it seems likely that influencetactics will have strongereffects on performanceassessments
than on measuresof extrinsic success.
Copyright? 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
INFLUENCEAND OUTCOMES
93
Source of performance assessment
Wortmanand Linsenmeier(1977) have arguedthat certaininfluence tactics are likely to be less effective in the workplacethan in an employmentinterview.Furthermore,they have suggested that differences in how individuals are evaluatedmight depend on the amount of backgroundinformationthe
rater has about the ratee. For example, in the workplace, raters typically have the opportunityto
observethe rateeover a long periodof time and acrossmany situations.As a result,it would be difficult
for the ratee to consistently employ influence tactics that might affect the evaluationof his or her performanceby a supervisor.In an interview,however,interviewersare able to observe applicantsfor only
a limited amountof time and in only one situation.Because interviewershave a more limited knowledge of applicants'backgrounds,their perceptionsmust be based primarilyon this limited period of
observation.Thus, when applicantschoose to employ influencetactics duringthe interview,it is likely
that they will be able to influence the ratings given to them by interviewers.As a result, we expect
influence tactics to have strongereffects on performanceassessments provided by interviewersthan
those providedby supervisors.
Research environment
The environmentin which the primarystudieswere conducted(laboratoryversus field) is also likely to
affect the strengthof the relationshipsbetween influence tactics and work outcomes. In a laboratory
environment,researchersare able to exert more controlover the situationand manipulateindependent
variablesmore precisely thanresearchersin field studies,where a numberof interveningvariablesmay
detractfrom the effectiveness of influence tactics. In addition,much as the interviewcontext provides
only a brief interactionbetween the influencerand the targetof influence,laboratoryexperimentsalso
tend to be brief encountersin which influence tactics are likely to have more impact than they might
over a longer period of time. As a result, we expect influence tactics to have strongereffects on work
outcomes in a laboratoryenvironmentthan in the field.
OrganizationalContext
Contextual Information About the Studies Included in the Meta-Analysis
The primarystudies included in our meta-analysiswere conducted between 1973 and 2000. The
majority (94 per cent) were conducted between 1980 and 2000. Twenty-seven studies appeared
in academicjournalsand four were unpublisheddoctoraldissertations.Only two studies were conducted outside the United States. Both of these studies were conductedin India.Finally, 29 studies
were cross-sectional in naturewhile two were longitudinal.
Method
Rulesfor inclusion
Primarystudies were includedin our analyses if they met the following conditions.First, an appropriate effect size statistic (e.g., correlationor d-value) must have been reported.If no effect size was
Copyright 2003 JohnWiley& Sons,Ltd.
?
J. Organiz.Behav.24, 89-106 (2003)
94
C. A. HIGGINSETAL.
reported,the informationnecessaryfor computingthe effect size (i.e., means and standarddeviations)
must have been available either in the study or from the study author(s).
Second, influence tactics must have been one of the following: assertiveness,coalitions, exchange,
ingratiation,rationality,upwardappeal, or self-promotion(however,because we found only two studies that examined the relationshipbetween coalitions and work outcomes, coalitions was subsequently droppedfrom our study). Several studies appearedto assess these constructsbut identified
the tactics by differentnames. As a result, it is possible that these studies actually examined different
constructs. In order to address this concern, three individuals familiar with the influence literature
independentlyevaluated each study included in our analysis. Each individual searched each study
for a definitionor descriptionof the tactics underexaminationor, when available, for the items used
to measureeach tactic. Then, each tactic was placed into one of eight categoriesrepresentingthe seven
tactics originally includedin the presentstudy and a category labelled 'other.'The lattercategorywas
to be used if the tactic in questiondid not fit the definitionof any of the tactics used in the presentstudy.
Following the independentevaluationof these studies, the three sets of evaluationswere comparedto
determine areas of discrepancy.There was disagreementin only one case and the three evaluators
discussed this discrepancyuntil agreementwas reached.
Third,following Ferrisand Judge's (1991) classificationof outcomes of influence behaviors,work
outcomes must have been a measureof performanceor competence,salary,or promotions.In addition,
measuresof salary recommendationor progressionand promotionrecommendationor promotability
were included. Again, based on Judge and Bretz (1994) all measuresof salary and promotionswere
combined to create the extrinsic success variable.
Identificationof studies
In an effort to obtain the broadestpossible sample of empirical studies relating influence tactics to
work outcome variables, we conducted a computerizedsearch of the PsychINFO database (19672000). Because the influence tactic constructis identifiedby a variety of names including impression
management(Tedeschi & Melburg, 1984), organizationalpolitics (Ferriset al., 1989) political influence behavior (Judge & Bretz, 1994), and influence tactics (Thacker & Wayne, 1995), multiple
searches were necessary. In addition, our dependent variable also consists of a number of different
facets: salary, promotions, and performanceassessments, both by supervisors and by interviewers.
Again, this requiredmultiple searches. These searches consisted of reading more than 300 abstracts
to identify studies that examined influence tactics and work outcomes. Of these approximately300
studies, 81 were identifiedfor possible inclusion in our meta-analysis.These 81 studies included 65
publishedjournalarticlesand 16 unpublisheddoctoraldissertations.Each of these 81 studies was then
obtained and read in order to ascertainwhether or not the study met the criteriafor inclusion in the
present study.
Based on the previously stated inclusion rules, 23 studies were retainedfor our analysis. The most
common reasonfor discardingstudies was a failureto reportcorrelations,d-values, or the information
necessaryto computeeffect sizes thatcould be used in our analyses. When studies failed to reportthis
information,the study authorswere contacted in an effort to obtain correlationsor d-values. If the
authorswere unable to provide that information,the study was eliminated from consideration.
A smallernumberof studies were excluded based on the definitionand/ormeasurementof influence
tactics. Studies eliminatedbased on definitionor measurementdifferencestypically combined several
influence tactics into one variableor one manipulation(e.g., video stimulus) making it impossible to
extract the effect of each influence tactic on the dependentvariableof interest.
Copyright? 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
INFLUENCEAND OUTCOMES
95
The next step in identifying possible studies consisted of a manual search of the reference lists of
those studies already included in our meta-analysis.The reference list search resulted in three additional studies that met the inclusion criteria.Finally, we contacted severalresearchersin the influence
tactic field who identifiedfive additionalstudies, which were subsequentlyincluded in the final analyses. All told, 31 studies were retainedfor analysis. These 31 studies contained37 independentsamples and reported98 correlationsbetween individualinfluence tactics and work outcomes.
Meta-analyticprocedures
The meta-analyticproceduresoutlined by Hunter and Schmidt (1990) were used in conducting our
analyses. When correctingcorrelationsindividually,as was done in the present study, the procedure
developed by Hunterand Schmidt stipulatesthat observed correlationsbe correctedfirstfor unreliability in both the predictorand criterionvariables,then the meta-analysisis conductedon the corrected
correlations.The correctionsfor unreliabilityare discussed in the following sections.
Correcting for unreliability in influence tactic measures
A majority(87 per cent) of the studies included in our meta-analysiscontainedinternalconsistency
reliabilities for the influence tactic measuresused in the study.When such a reliability estimate was
available, it was used in all subsequentanalyses of that study.When no reliabilitywas reportedfor a
given measure,an estimatedreliabilitywas calculatedusing the averageof the reliabilitiesreportedin
other studies in our sample that assessed the same influencetactic. In the presentstudy,four studies of
ingratiationand one study of self-promotionfailed to reportreliability estimates. The averagecoefficient alphareliabilitiesthatwere computedand used for these studies were a = 0.71 (ingratiation)and
a = 0.80 (self-promotion).
Typically, if influence tactics were used in a selection context where the criterionwas job performance, one would not correct influence tactics for reliability.Because this is not the context of the
present study, however, it is reasonable to correct influence tactics for unreliabilityeven when the
criterion is job performance.
Correcting for unreliability in work outcome measures
Although most studies that included performanceassessments as a dependentvariableincluded estimates of internalconsistency reliability,these estimates were not used in our analyses. Because we
typically seek to generalize,not to one specific rater,but to an equally knowledgeablerater,the appropriate correctionto use is inter-raterreliability.Therefore,for studies that included supervisorperformance ratingswe used the inter-raterreliability estimate of 0.52 reportedby Viswesvaran,Ones, and
Schmidt (1996). This meta-analyticallyderivedestimate,based on a sample of 14 650 subjects,represents the most thoroughestimate to date of the mean populationestimate of supervisoryperformance
ratings.For studies that included performanceassessmentsby interviewers,we correctedfor unreliability using the inter-raterreliabilityestimate of 0.70 reportedby Conway,Jako,and Goodman(1995)
as the mean populationestimateof inter-raterreliabilityin interviewsettings. Again, this estimatewas
derived throughmeta-analytictechniques and is the most accurateestimate available for inter-rater
reliability in an interview setting.
The remainingwork outcome measures, salary and promotions,were typically obtainedthrougha
one-time searchof historicalrecordsby one individual.When this occurs there is no associatedinternal consistency reliability or inter-raterreliability that could be used in correcting the observed
correlationsfor unreliability.Schmidt and Hunter(1996) addressedthe issue of unreliabilityin such
objective measures and concluded that it is appropriateto assume perfect reliability with the
Copyright? 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
96
C. A. HIGGINSETAL.
understandingthat all resultingcoefficients will be downwardlybiased to some unknownextent. As a
result,the reliabilityfor all measuresof salaryand promotionswas coded as 1.0. Therefore,all resulting correlationcoefficients are conservativeestimates of the true relationship.
Creating composite measures
Severalstudiesreportedmultiplemeasuresof eitherinfluencetactics, work outcomevariables,or both.
When such studies were encountered, equally weighted composite measures were constructed as
recommendedby Hunterand Schmidt (1990). That is, if a study reportedcorrelationsbetween ingratiation and three differentmeasuresof performancea composite was computedbased on the average
correlationbetween ingratiationand the performancemeasures,correctedfor the correlationamong
the three measuresof performance(see Hunter& Schmidt, 1990, p. 457). The purposeof such a composite correlationis to estimatethe correlationbetween ingratiationand overallperformanceas if such
an overall correlationhad been reported.By computingcomposites from these samples, the independence of the sample is maintained.In addition,creatinga composite correlationgenerallyresults in a
more constructvalid measureof the variableof interest(Ones & Viswesvaran, 1996).
Results
Overall analysis
The results of the overall meta-analysesare presentedin Table 1. The first set of analyses was conductedwith each influencetactic as an independentvariable.All individualwork outcomes were combined into a single dependentvariablefor overall analyses. Results of these analyses were quite varied
with mean weighted, uncorrectedcorrelationsranging from -0.01 (exchange) to 0.18 (rationality).
After correctingfor predictorand criterionunreliability,estimatedpopulationmean correlationsranged from -0.03 (exchange) to 0.26 (rationality).In addition,the varianceaccounted for by artifacts
rangedfrom 6.9 per cent (self-promotion)to 73.5 per cent (exchange). These results suggest substantial variabilityin the effectiveness of differentinfluence tactics. However, it is importantto note that
the upwardappeal analyses were based on only four studies. Thus, results of this analysis are tenuous
and should be interpretedwith caution.
In order to assess variability in the individual, corrected true score correlations, 95 per cent
credibility intervals(CVs), which provide informationconcerningthe dispersion of individual study
correlations,were computedfor each set of relationships.The credibilityintervalswere generallyquite
Table1. Theeffectof influencetacticson overallworkoutcomes
Influencetactic k
Meanr
SDr
Meanp
Ingratiation 50 6065
Self-promotion20 3587
0.17
0.02
0.17
0.28
0.23
0.01
SDP
0.26
0.35
n
%Variance 95%CV
20.0
6.9
(-0.28, 0.74)
(-0.68,0.70)
95%CI
(0.20,0.26)
(-0.03,0.05)
Rationality
9
1454
0.18
0.16
0.26
0.26
13.7
(-0.25, 0.77)
9
1487
0.00
0.16
-0.02
0.23
18.7
(0.20, 0.32)
Assertiveness
(-0.47, 0.43)
(-0.08, 0.04)
Exchange
Upwardappeal
6
4
1041 -0.01
846 0.04
0.05
0.15
-0.03
0.05
0.07
0.22
73.5
16.9
(-0.17,0.11)
(-0.38, 0.48)
(-0.10,0.04)
(-0.03, 0.13)
deviationof
Notes:k= numberof correlations;
n = totalsamplesize;meanr = averageuncorrected
correlation;
SDr= Standard
deviationof correctedcorrelation;
uncorrectedcorrelation;mean p=average correctedcorrelation;
SD,=-standard
= lowerandupperlimitsof 95%credibility
= percentvarianceaccountedforby artifacts;
%variance
interval;
95 percentCV
95%CI= lowerandupperlimitsof 95%confidenceinterval.
Copyright D 2003 John Wiley & Sons, Ltd.
J. Organiz.Behav. 24, 89-106 (2003)
INFLUENCEAND OUTCOMES
97
large ranging from 0.28 (exchange) to 1.38 (self-promotion)suggesting substantialvariabilityin the
individualcorrelationsacross studies.
Additionalanalysesproduced95 per cent confidenceintervals(CIs) for each influencetactic. Unlike
credibility intervals,confidence intervalsprovide informationregardingvariabilityaroundestimated
mean populationvalues. Whereas CVs provide the range across which individual study correlations
are dispersed,CIs provide a measureof confidence concerningthe estimatedmean true correlationin
the population.Confidenceintervalsfor ingratiationand rationalityexcluded zero suggesting that the
average relationshipsbetween work outcomes and each of these influence tactics is non-zero.
As indicatedpreviously,our overall meta-analysessuggested that artifactsfailed to accountfor 100
per cent of the variancein any of the tested relationships.In addition,Q-tests were significantfor all
sets of effects sizes except exchange. This suggests that the remainingsets of effect sizes contained
significantamountsof unexplainedvariabilityacross studies.As a result additionalanalyses were conducted in an attemptto explain additionalbetween studies variance.
Workoutcomes
The firstset of additionalanalyses examinedthe independenteffects of influencetactics on each work
outcome. Results of these analyses are presentedin Table 2 and highlights are discussed below.
Ingratiation and work outcomes
The analysis of ingratiationby work outcome measureindicatedthatingratiationhad a generallypositive effect on both work outcomes. The strongesteffect was found for performanceassessmentswith a
Table2. Effectof influencetacticson specificworkoutcomemeasures
Measure
k
n
Mean Mean
r
p
%Variance 95%CV
95%CI
Significant
differences
Ingratiation
(a) Performance
assessments
(b) Extrinsic
27
3090
0.23
0.35
17.8
(-0.24, 0.94)
(0.31,0.39)
b
23
2975
0.09
0.11
47.2
(-0.13, 0.35)
(0.06, 0.16)
a
11
1774
0.03
0.01
4.4
9
1813
0.01
0.01
100.0
(0.01, 0.01)
(-0.05, 0.07)
4
547
0.32
0.50
13.8
(-0.01, 1.1)
(0.41,0.59)
b
5
907
0.10
0.12
92.5
(0.06, 0.18)
(0.04, 0.20)
a
4
547 -0.12
-0.19
20.5
(-0.72, 34)
(-0.29, -0.09)
b
5
940
32.9
(-0.17, 0.33)
(0.00,0.16)
a
success
Self-promotion
(a) Performance
assessments
(b) Extrinsic
(-0.95, 0.97) (-0.04, 0.06)
success
Rationality
(a) Performance
assessments
(b) Extrinsic
success
Assertiveness
(a) Performance
assessments
(b) Extrinsic
0.07
0.08
success
Notes:k = numberof correlations;
meanp = averagecorrected
n = totalsamplesize;meanr = averageuncorrected
correlation;
%variance=percentvarianceaccountedfor by artifacts;
95%CV= lowerandupperlimitsof 95%credibility
correlation;
interval;95%CI= lowerandupperlimitsof 95%confidenceinterval.
Lettersin theright-hand
columndenotemeansthataresignificantly
differentfromeachotheratthe0.01level.Meanswerenot
comparedacrossinfluencetactics.
Copyright( 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
98
C. A. HIGGINSETAL.
mean populationcorrelationof 0.35. This effect is larger (p = 0.35 versus p = 0.05) but in the same
direction as the result of similar analyses reportedby Gordon(1996).
To determinewhetherthe effects of ingratiationon performanceassessmentsand extrinsic success
were statisticallydifferentfrom each other, pairwise comparisonswere conductedusing Z-tests suggestedby Quifiones,Ford,andTeachout(1995). Resultsof these tests indicatedthatingratiationdid have
a significantlystrongereffect on performanceassessmentsthanon extrinsicsuccess (z = 7.71, p < 0.01).
Self-promotion and work outcomes
Self-promotionwas found to have only a weak effect on performanceassessments and extrinsic success. Although mean populationcorrelationswere slightly positive in both instances, both confidence
intervals contained zero, leaving the possibility that the true effect of self-promotionon these work
outcomes is zero.
Rationality and work outcomes
As with ingratiation,more detailed analyses of rationalityand work outcome measures suggested a
positive influence of rationalityon both performanceassessments and extrinsic success. Rationality
also had a significantly stronger effect on performance assessments than on extrinsic success
(z = 11.93, p < 0.01). However,the results of these analyses should be viewed with cautionas no subgroup contained more than five individualstudies.
Assertiveness and work outcomes
The analysis of assertiveness and work outcomes provided interestingresults. Whereas a negative
effect was found for performanceassessments,a positive effect was found for extrinsic success resulting in a significantdifferencebetween these two relationships(z = 7.94, p < 0.01). Althoughbased on
a limited numberof studies, these results are cause for speculationregardingthe differentialeffect of
assertivenesson two seemingly related work outcomes.
Source of performanceassessment
Anothervariablewe expected to affect the relationshipbetween influence tactics and work outcomes
was the source of the performanceassessments.Thus, our next set of analyses focused on the relationship between influencetactics and performanceassessmentsgiven by supervisorsversus performance
assessments providedby interviewers.As the results in Table 3 show, ingratiationand self-promotion
had significantlystrongereffects on performanceassessmentsprovidedby interviewersthanthose provided by supervisors (ingratiation:z = 7.90, p < 0.01; self-promotion:z = 25.31, p < 0.01). In fact,
whereas self-promotion had a strong, positive effect on interviewers' performanceassessments, a
strong, negative effect was found for supervisorassessments. However, we must urge caution in the
interpretationof these latter results as only three studies were used in computing the meta-analytic
estimate of the relationshipbetween self-promotion and interviewerperformanceassessments. As
such, a single study may have unduly influencedthe resultingestimate making these results, at best,
speculative. In orderto better understandthis relationship,furtherstudy is necessary.
Research environment
Table 4 presents the results of the final analysis. This analysis grouped studies by researchenvironment: laboratoryversus field. However,the only influencetactic for which multiple studies have been
conductedin both the laboratoryandin the field is ingratiation.In fact, of all the otherinfluencetactics,
Copyright? 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
INFLUENCEAND OUTCOMES
99
Table 3. Effect of performanceassessment source on the relationshipbetween influencetactics and performance
assessments
Measure
k
Ingratiation
(a) Supervisorperformance
assessments
(b) Interviewerperformance
assessments
Self-promotion
(a) Supervisorperformance
assessments
(b) Interviewerperformance
assessments
n
Mean
r
20 2299
7
8
3
791
0.17
0.26
19.9
0.41
0.60
1215 -0.18
559
Mean % Variance
p
0.49
-0.25
95% CV
(-0.33,0.85)
95% CI
Significant
differences
(0.21,0.31)
b
58.1
(0.42, 0.78) (0.53, 0.67)
a
23.9
(-0.64, 0.14) (-0.31, -0.19)
b
2.0
(-0.34, 1.50) (0.51, 0.65)
a
0.58
Notes:k= numberof correlations;
n = totalsamplesize;meanr = averageuncorrected
meanp = averagecorrected
correlation;
= percentvarianceaccountedfor by artifacts;95%CV= lowerandupperlimitsof 95%credibility
% variance
correlation;
interval;95%CI= lowerandupperlimitsof 95%confidenceinterval.
Lettersin right-hand
columndenotemeansthataresignificantly
differentfromeachotherat the 0.01 level.Meanswerenot
acrossinfluencetactics.
compared
only self-promotionand rationalityhave been included in even one laboratorystudy with work outcomes as the dependentvariable.
In orderto avoid confoundingthe resultsby using a compositedependentvariable(i.e., one thatconsistedof a combinationof performanceassessmentsandextrinsicsuccess),two analyseswere conducted.
The first analysis examinedthe effect of ingratiationon performanceassessmentsin laboratoryversus
field environments.The secondexaminedthe effect of ingratiationon extrinsicsuccess in laboratoryversus field environments.In each instance,significantlystrongereffects were foundin the laboratorythanin
the field (performanceassessments:z= 3.78, p < 0.01; extrinsicsuccess: z = 4.77, p < 0.01).
Discussion
Drawing from an older literaturein social psychology, organizationalresearchersin the last 20 years
have linked influence tactics to a numberof criteriathat reflect work outcomes. The breadthof these
Table4. Effectof researchenvironment
on the relationship
andworkoutcomes
betweeningratiation
Method
k
n
Mean Mean %Variance 95%CV 95%CI
Significant
r
p
differences
Performance
assessments
10 656
(a) Laboratory
17 2434
(b) Field
Extrinsicsuccess
(a) Laboratory
(b) Field
0.33
0.21
0.48
0.31
19.2
17.8
654
0.25
0.29
52.2
13 2321
0.05
0.06
84.8
10
(-0.21, 1.17) (0.40,0.56)
(-0.24,0.86) (0.26,0.36)
(0.04, 0.55)
(0.21, 0.37)
(-0.02,0.14) (0.01,0.11)
b
a
b
a
Notes:k= numberof correlations;
n = totalsamplesize;meanr = averageuncorrected
meanp = averagecorrected
correlation;
%variance= percentvarianceaccountedforby artifacts;95%CV= lowerandupperlimitsof 95%credibility
correlation;
interval;95%CI= lowerandupperlimitsof 95%confidenceinterval.
Lettersin theright-hand
columndenotemeansthataresignificantly
differentfromeachotherat the0.01level.Meanswerenot
comparedacrossworkoutcomes.
Copyright(D 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
100
C. A. HIGGINSETAL.
studies is impressive;researchershave studied a wide range of influencetactics, and relatedthese tactics to many differentwork outcomes. At the same time, this breadthhas had its drawbacks.The literaturehas providedremarkablylittle consensus on the relativeeffectiveness of these influencetactics.
Therefore,the presentstudy sought to addressthe issue of which influencetactics are most effective in
helping one obtain desirablework outcomes.
Results show that two influence tactics (i.e., ingratiationand rationality)displayed mean non-zero
relationshipswith work outcomes. Thus, individuals who exhibit ingratiatorybehaviors, and who
use logic anddatato justify theirrequests,appearto have a greaterchance of succeedingin theircareers
than individualswho use these tactics to a lesser degree. However, while these influence tactics displayed the strongestcorrelationswith work outcomes, and we can be highly confidentthatthe average
correlationis non-zero, the individualcorrelationswere highly variableacross studies. Thus, it was
importantto investigatethe relationshipsbetween influencetacticsand workoutcomesin greaterdetail.
In fact, the resultsof the additionalanalyses were perhapsas interestingand importantas the results
from the overall analyses. Ingratiationand rationalityboth correlatedmuch more strongly with performance assessments than with extrinsic success. Why were these differences by criteriaobserved?
One possible reason for these differential relationshipsis that performanceassessments are more
directlyinfluencedby employee behaviorthanare salaryor promotions,the two componentsof extrinsic success. If influence tactics affect salary or promotions,they likely do so throughtheir effects on
performanceassessmentssince such assessmentsare a primarycomponentof many salaryandpromotion decisions. Thus, influence tactics probablymore proximally influence performanceassessments,
and their effect on salary and promotionsis probablyonly indirect.Another,related, factor that may
explain differential relations by criteria is the issue of supervisorypower and influence. Typically,
upward influence tactics are designed to motivate supervisorsto produce the outcomes desired by
the subordinate.Performanceassessments are usually directly within supervisors'control. However,
salary and promotionsare less clearly within the supervisor'scontrol and likely are more affected by
externalconditionssuch as the financialconditionof the organizationand the availabilityof open positions to which one can be promoted.Thus, ingratiationandrationalitymay have a strongerrelationship
with performanceassessmentsbecause these outcomes are within the direct control of the typical target of these tactics (i.e., the supervisor)whereasthe targettypically has much less control over salary
and promotions.
On a similar note, assertivenessalso had differingrelationshipswith performanceassessments and
extrinsic success. On the one hand, assertivenesshad a moderatelystrong, negative relationshipwith
performanceassessments. On the other hand, assertivenesshad a positive relationshipwith extrinsic
success. One might typically assume thatan influencetactic would affect both of these workoutcomes
in the same manner(i.e., eitherpositively or negatively). However,resultsof the presentstudy suggest
that this is not always true. One explanationmay be thatthose who use assertivenessare more aggressive in seeking out or asking for pay raises and promotions.In so doing, those individualsmay receive
rewardsthatoutpacetheiractualperformanceas reflectedin theirperformanceratings.Althoughthese
results are based on a relatively small numberof studies (k = 4 and 5, respectively), future research
should examine these relationships in more detail in an effort to elucidate the different processes
throughwhich assertivenessaffects various work outcomes.
Anotherimportantfactor was the source of performanceassessments.When influence tactics were
studiedin the context of employmentinterviews,the effect sizes were quite large. The intervieweffect
is particularlynotablefor self-promotion.It appearsthatself-promotionaids in the receiptof favorable
interview ratings, but such tactics seem to backfirein obtaining favorable performanceassessments
from supervisors.This latterresultis consistentwith previousworkby Gordon(1996) and makes sense
if one considers the natureof self-promotion.Jones and Pittman(1982) suggested that self-promotion
is likely to be more successful when claims of competence are difficult to verify. On the other hand,
Copyright ) 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
INFLUENCEAND OUTCOMES
101
when claims of competence can be easily refuted the chances of achieving success are likely to be
diminished. Therefore,when one self-promotesin an interview,the likelihood of establishing a perception of competence is good because the interviewer has little ability to verify. Conversely, the
supervisoris probably in a position to gauge the veridicality of self-promotion,because compared
to the interviewer,the supervisorhas a greaterability to recognize self-promotionfor what it is, rather
than as an accurate description of one's accomplishments. However, a note of caution should be
sounded here. Only three studies that investigated self-promotionin the interview were included in
this meta-analysis.Although a meta-analysisbased on only three studies is generally more interpretable thanresultsfrom a single study (Hunter& Schmidt, 1990), one can be less confidentthatwe have
close estimates of populationvalues than if the numberof studies was larger.
Finally,results of the presentstudy suggest that ingratiationhas a much strongereffect on work outcomes in the laboratorythanin the field. This is consistentwith resultsreportedpreviouslyby Gordon
(1996) and is not surprisingdue to the ability of researchersto develop strongexperimentalmanipulations of ingratiationin the absence of many of the contextualfactors presentin field studies that work
to mitigatethe effects of influencetactics (i.e., accountability,historicalrelationshipbetween the target
and the influencer,etc.).
Limitationsand contributions
The presentstudyhas severallimitationsthatneed to be noted. First,due to the relativelylarge number
of influence tactics that were included in the present study and the relatively small numberof studies
that fit the criteriafor some analyses, we were limited in our investigationof factors that might affect
the relationshipsbetween influencetactics and work outcomes to specific work outcome measures,the
source of performanceassessments, and researchenvironment.Given that the correlationof the two
most effective influencetactics (i.e., ingratiationandrationality)with work outcomes was highly variable across studies, it would have been informativeto conduct more extensive analyses. As studies in
this area continue to cumulate,it would be importantfor futureresearchto investigateadditionalfactors that may influence the relationshipbetween influence tactics and work outcomes.
Related to this issue is the fact that some influence tactics have not received much attentionin the
researchliterature.Therefore,as noted above, the numberof studies included in some analyses was
small. For example, despite the fact that rationalityappearsto have a significanteffect on work outcomes, we identifiedfewer than 10 studiesthatempiricallytested the effects of rationalityin the workplace. Furthermore,exchange and upwardappealhave been the focus of even fewer studies. Although
ingratiationand self-promotionhave received considerableattention,it is importantfor futureresearch
to examine other tactics in order to gain a comprehensiveunderstandingof the effects of influence
tactics in organizations.
A thirdlimitationis thatthis studyprovideda relativelydirectview of the effectiveness of influence
tactics. In most models of influence tactic use, a numberof contextualfactors are taken into account,
such as political norms (Porter,Allen, & Angle, 1981), affect or liking (Ferris& Judge, 1991), characteristics of the influence target (Mowday, 1978), and direction of influence (Yukl & Falbe, 1990).
Unfortunately,as is the case with most meta-analyses,it is not possible to code these variablesat the
study level. Although this is a limitation of our study, it is also an importantdirection for future
research.
The above limitationsnotwithstanding,we believe this papermakes an importantcontributionto the
influence tactic literature.Despite increasedinterestin the last 20 years, researchhas not appearedto
successfully answer one of the most fundamentalquestions;that is, which influence tactics are most
effective in obtainingpositive work outcomes? Ferrisand Judge (1991) provideda qualitativereview
Copyright? 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
102
C. A. HIGGINSETAL.
of the relationshipbetween influence tactics and work outcomes. Yet, in examining their review, it is
difficult to come away with a clear understandingof which influence tactics appearto be most (and
least) effective in obtainingdesirablework outcomes. Indeed,as FerrisandJudgenoted with respectto
performanceevaluations, 'Clearly furtherresearchis needed in this area to more precisely delineate
the differentialeffectiveness of a broadarrayof influenceattempts'(p. 464). Ferrisand Judgewent on
to call for more researchon the ' ... differentialeffectiveness of particularforms of political influence
in different.., .contexts' (p. 470). By revealingthatingratiationand rationalitydisplay mean, non-zero
relationshipswith work outcomes, that ingratiationand self-promotionare stronglyrelatedto success
in employmentinterviews,and that self-promotionand assertivenessmay backfirein some situations,
hopefully our work has added some clarity to this area.
Implicationsandfuture research
Individualswho desire to be successful in their careerswould benefit from knowing that some influence tactics, especially ingratiationand rationality,appearto be effective means of influencingothers.
In addition,some of the relationshipsinvestigatedin the currentstudy are contextually sensitive. For
example, ingratiationand self-promotionappearto work particularlywell in the interview and, in the
case of self-promotion,it appearsto backfirein performanceevaluationsprovidedby supervisors.It
seems then, thatindividualsseeking to become more politically astutecould be informedby the results
of this study.
Although we were unable to directly test the effectiveness of combinationsof tactics in the current
study,combining our results with those reportedby Falbe and Yukl (1992) allows us to speculatethat
certain combinationsof tactics may be particularlysuccessful in obtaining desirablework outcomes.
To illustrate,Falbe and Yukl reportedthat combining two soft tactics or a soft tactic with rationality
would lead to the greatest likelihood of success. Results of the currentstudy indirectly supportthis
assertion as ingratiation(a soft tactic) and rationalityconsistently had the strongestpositive relationships with work outcomes. Thus, it appearslikely that combining ingratiationand rationalitywould
provideone with a good chance of obtainingpositive work outcomes. However,as we found no studies
examining the effects of influence tactic combinationson work outcomes, this is an importantavenue
of futureresearch.
There are a numberof importantquestions that cannot be answered in a quantitativereview, and
have not yet been addressed in the literature.First, though interest in influence tactics is growing,
one of the issues fundamentalto any field of researchhas yet to be fully addressedin the influence
literature.That issue is the measurementof influencetactics. Althoughresearchershave been studying
influence tactics for more than 30 years, there is still no unified conceptualizationof what influence
tactics are or how they should be measured.Proof of this exists in the primarystudies included in the
presentmeta-analysis.The 31 studiesrepresentedhereused at least nine differentinstrumentsfor measuring influence tactics. While having multiple instrumentsfor measuring a given construct is not
inherentlywrong, it is importantto know whether each of these instrumentsis measuringthe same
underlyingfactors. In order to do this, we must know what those underlyingfactors are. Therefore,
one importantavenue for future research is the development of a comprehensive frameworkthat
clearly defines what influencetactics are and what they are not. We shouldnote, however,thatdiversity
in measures did not undermineour ability to find non-zero effects with respect to several influence
behaviors.
Anotherareafor furtherresearchis the role of two new influencetactics recentlyaddedto the Kipnis
et al. (1980) typology by Yukl and Falbe (1990); that is, inspirational appeal and consultation.
Researchby Yukland associates has found thatinspirationalappealis among the most effective tactics
Copyright? 2003 JohnWiley& Sons,Ltd.
J. Organiz.Behav.24, 89-106 (2003)
INFLUENCE
ANDOUTCOMES 103
in gaining cooperationfrom others (Yukl & Falbe, 1996). However, because we found no study that
linked the use of this tactic to work outcomes, we were not able to include it in our review.
Next, the directionof the influenceattempt(i.e., upwardversus lateralversusdownward)is a potential moderatorof the relationshipsexamined in the currentstudy. However, we were able to identify
only one study of lateraland downwardinfluence attempts(Yukl& Tracey, 1992) thatinvestigatedthe
relationshipbetween influence tactics and work outcomes. Despite the attentiongiven to this issue in
recentyears (e.g., Yukl& Falbe, 1990) it appearsthatfew have attemptedto empiricallytest this aspect
of the influencerelationship.It seems likely that differinglevels of success may be realizedwhen using
certaintactics with subordinates,peers, and supervisors.Forexample, whereasthe presentstudyfound
self-promotionto have a negative effect on performanceassessments made by supervisors,it is possible thatthe use of self-promotiontowardsubordinatesmay resultin thatgroupprovidingmore positive assessments of managerialperformance.One explanation for such a result could be that the
subordinategroup does not have the means to verify claims made by the influencer,whereas a supervisor is in a much better position to determine the veracity of such claims, especially when those
claims are work-related.Therefore,future researchthat examines the moderatingeffect of direction
on the influence tactic/work outcome relationship, stands to make a significant contributionto the
influence literature.
In addition to the direction of the influence attempt,the level, or amount, of influence may also
affect the outcome. Whereaspreviousresearchhas suggestedthattoo much ingratiationmay adversely
affect the outcome of an influence attempt(Baron, 1986), it is unclearif such non-lineareffects exist
with other influence tactics. It seems plausible to expect that the effects of self-promotionmight be
non-linear,such that a moderateamount may be effective in some situations while too much or too
little may be ineffective. On the otherhand,it seems one would be hardpressed to use too muchrationality. Thus, futureresearchshould examine the possibility of non-linearrelationshipsbetween influence tactics and work outcomes.
Finally, future researchneeds to investigate the style or political skill of the influencer.The style
used by influencers likely contributessubstantiallyto the effectiveness of influence tactics because
of the influencer'sability to disguise ulteriormotives of manipulation,and instead contributeto positive motives and intentionsbeing perceived.Recent researchhas labelled such style considerationsas
'political skill,' and future work needs to examine this constructin more detail (Ferris et al., 2000;
Ferris, Berkson, Kaplan, Gilmore, Buckley, Hochwarter, & Witt, 1999-paper presented at the
Academy of Management59th Annual National Meeting, Chicago).
Acknowledgement
The authorsthank Frank Schmidt for helpful comments that contributedto the preparationof this
article.
Author biographies
Chad Higgins is Assistant Professorof HumanResource Managementand OrganizationalBehavior,
Departmentof Managementand Organization,University of Washington.Chad holds a Bachelor of
Copyright? 2003 John Wiley & Sons, Ltd.
J. Organiz.Behav. 24, 89-106 (2003)
104
C. A. HIGGINSETAL.
Business Administration degree from the University of Nebraska and received a PhD in Business
Administration from the University of Iowa. Chad's research interests are in the areas of staffing,
job search behaviors, person-organization fit, and impression management.
Gerald R. Ferris is the Francis Eppes Professor of Management and Professor of Psychology at
Florida State University. Ferris received a PhD in Business Administration from the University of Illinois at Urbana-Champaign. He has research interests in the areas of social influence processes in
human resources systems, and the role of reputation in organizations.
Tim Judge is the Matherly-McKethan Eminent Scholar, Department of Management, Warrington
College of Business, University of Florida. Tim holds a Bachelor of Business Administration degree
from the University of Iowa, and Master's and Doctoral degrees from the University of Illinois. Tim's
research interests are in the areas of personality, leadership and influence behaviors, staffing, and job
attitudes.
References
References marked with an asterisk (*) indicate studies included in the meta-analysis.
*Baron,R. A. (1986). Self-presentationin job interviews:when there can be too much of a good thing. Journal
of Applied Social Psychology, 16, 16-28.
*Baskett, G. D. (1973). Interview decisions as determinedby competency and attitude similarity.Journal of
Applied Psychology, 57, 343-345.
Cleveland, J. N., & Murphy,K. R. (1992). Analyzing performanceappraisalas goal-directedbehavior.In G. R.
Ferris, & K. M. Rowland (Eds.), Research in personnel and humanresourcesmanagement(Vol. 10, pp. 121185). Greenwich,CT: JAI Press.
Conway, J. M., Jako, R. A., & Goodman, D. F. (1995). A meta-analysisof interraterand internal consistency
reliability of selection interviews.Journal of Applied Psychology, 80, 565-579.
Cooper, W. H., Graham,W. J., & Dyke, L. S. (1993). Tournamentplayers. In G. R. Ferris (Ed.), Research in
personnel and humanresources management(Vol. 11, pp. 83-132). Greenwich,CT: JAI Press.
*Deluga, R. J., & Perry,J. T. (1994). The role of subordinateperformanceand ingratiationin leader-member
exchanges. Group & OrganizationManagement,19, 67-86.
*Dulebohn,J. H., & Ferris,G. R. (1999). The role of influencetactics in perceptionsof performanceevaluations'
fairness.Academyof ManagementJournal, 42, 288-303.
*Eastman,K. K. (1994). In the eyes of the beholder:an attributionalapproachto ingratiationand organizational
citizenship behavior.Academyof ManagementJournal, 37, 1379-1391.
Falbe,C. M., & Yukl,G. (1992). Consequencesfor managersof using single influencetactics andcombinationsof
tactics. Academyof ManagementJournal, 35, 638-652.
Ferris,G. R., & Frink,D. D. (1997). Influencetactics in the performanceevaluationprocess. In L. H. Peters,C. R.
Greer, & S. A. Youngblood(Eds.), The Blackwell encyclopedic dictionary of human resource management
(pp. 165-167). Oxford, UK: Blackwell Publishers.
Ferris,G. R., & Judge, T. A. (1991). Personnel/humanresourcesmanagement:a political influence perspective.
Journal of Management,17, 447-488.
Ferris,G. R., Fedor, D. B., & King, T. R. (1994). A political conceptualizationof managerialbehavior.Human
Resource ManagementReview,4, 1-34.
*Ferris, G. R., Judge, T. A., Rowland, K. M., & Fitzgibbons, D. E. (1994). Subordinateinfluence and the
performanceevaluationprocess:test of a model. OrganizationalBehaviorand HumanDecision Processes, 58,
101-135.
Ferris, G. R., Perrewd,P. L., Anthony,W. P., & Gilmore, D. C. (2000). Political skill at work. Organizational
Dynamics, 28, 25-37.
*Fodor,E. M. (1974). Disparagementby a subordinateas an influence on the use of power. Journal of Applied
Psychology, 59, 652-655.
Copyright ) 2003 John Wiley & Sons, Ltd.
J. Organiz.Behav. 24, 89-106 (2003)
ANDOUTCOMES 105
INFLUENCE
Frink,D. D., & Ferris,G. R. (1998). Accountability,impressionmanagement,and goal settingin the performance
evaluationprocess. HumanRelations, 51, 1259-1283.
Gardner,W. L., & Martinko,M. J. (1988). Impressionmanagementin organizations.Journalof Management,14,
321-338.
Gerhart,B., & Rynes, S. L. (1991). Determinantsand consequences of salary negotiationsby male and female
MBA graduates.Journal of Applied Psychology, 76, 256-262.
*Gilmore,D. C., & Ferris,G. R. (1989). The effects of applicantimpressionmanagementtactics on interviewer
judgments. Journal of Management,15, 557-564.
Gilmore,D. C., Stevens, C. K., Harrell-Cook,G., & Ferris,G. R. (1999). Impressionmanagementtactics. In R. W.
Eder, & M. M. Harris(Eds.), The employmentinterviewhandbook(pp. 321-336). ThousandOaks, CA: Sage
Publications.
Godfrey,D. K., Jones, E. E., & Lord, C. G. (1986). Self-promotionis not ingratiation.Journalof Personalityand
Social Psychology, 50, 106-115.
Goffman, E. (1955). On face work: an analysis of ritualelements in social interactions.Psychiatry,22, 225-237.
Gordon,R. A. (1996). Impactof ingratiationon judgmentsandevaluations:a meta-analyticinvestigation.Journal
of Personalityand Social Psychology, 71, 54-70.
*Howard,J. L., & Ferris,G. R. (1996). The employment interview context: social and situationalinfluences on
interviewerdecisions. Journal of Applied Social Psychology, 26, 112-136.
Huber,V. L., Latham,G. P., & Locke, E. A. (1989). The managementof impressionsthroughgoal setting. In R. A.
Giacalone, & P. Rosenfeld (Eds.), Impressionmanagementin the organization (pp. 203-217). Hillsdale, NJ:
LawrenceErlbaum.
Hunter,J. E., & Schmidt,F. L. (1990). Methods of meta-analysiscorrectingerror and bias in researchfindings.
NewburyPark,CA: Sage Publications.
Jones, E. E. (1964). Ingratiation.New York:Appleton-Century-Crofts.
Jones, E. E. (1990). Interpersonalperception. New York:W.H. Freeman.
Jones, E. E., & Pittman, T. S. (1982). Towarda general theory of strategic self presentation.In J. Suls (Ed.),
Psychological perspectives on the self (pp. 231-262). Hillsdale, NJ: LawrenceErlbaum.
*Judge,T. A., & Bretz, R. D. (1994). Political influencebehaviorand careersuccess. Journalof Management,20,
43-66.
Kipnis, D., & Schmidt, S. M. (1982). Profile of organizational influence strategies. San Diego: University
Associates.
Kipnis, D., & Schmidt, S. M. (1988). Upwardinfluence styles: relationshipwith performanceevaluations,salary,
and stress.AdministrativeScience Quarterly,33, 528-542.
*Kipnis, D., & Vanderveer,R. (1971). Ingratiationand the use of power. Journal of Personality and Social
Psychology, 17, 280-286.
influence tactics: explorationsin getting
Kipnis, D., Schmidt, S. M., & Wilkinson,I. (1980). Intra-organizational
one's way. Journal of Applied Psychology, 65, 440-452.
Lawler, E. E., III. (1966). Managers' attitudestoward how their pay is and should be determined.Journal of
Applied Psychology, 50, 273-279.
Leary, M. R. (1995). Self presentation: Impression management and interpersonal behavior. Boulder, CO:
Westview Press.
Leary,M. R., & Kowalski,R. M. (1990). Impressionmanagement:a literaturereview and two-componentmodel.
Psychological Bulletin, 107, 34-47.
*Martin,D. C. (1987). Factorsinfluencingpay decisions: balancingmanagerialvulnerabilities.HumanRelations,
40, 417-430.
*Nowicki, C. E. (1983). Impression managementin the employment interview (Doctoral dissertation,Purdue
University, 1983). Dissertation AbstractsInternational,44, 3969.
Mowday,R. T. (1978). The exercise of upwardinfluence in organizations.AdministrativeScience Quarterly,23,
137-156.
Ones, D. S., & Viswesvaran, C. (1996). Bandwidth-fidelitydilemma in personality measures for personnel
selection. Journal of OrganizationalBehavior 17, 609-626.
*Orpen, C. (1996). The effects of ingratiationand self-promotiontactics on employee career success. Social
Behavior and Personality,24, 213-214.
*Pandey, J., & Kakkar, S. (1982). Supervisors' affect: attraction and positive evaluation as a function of
enhancementof others. Psychological Reports, 50, 479-486.
*Pandey,J., & Singh, P. (1987). Effects of machiavellianism,other-enhancement,and power-positionon affect,
power feeling, and evaluationof ingratiator.The Journal of Psychology, 121, 287-300.
Copyright? 2003 John Wiley & Sons, Ltd.
J. Organiz.Behav. 24, 89-106 (2003)
106
C. A. HIGGINSETAL.
*Peters, L. H., & Terborg,J. R. (1975). The effects of temporal placement of unfavorableinformationand of
attitude similarity on personnel selection decisions. OrganizationalBehavior and Human Performance,13,
279-293.
Porter,L. W., Allen, R. W., & Angle, H. L. (1981). The politics of upwardinfluence in organizations.In L. L.
Cummings, & B. M. Staw (Eds.), Research in organizational behavior (pp. 109-149). Greenwich, CT: JAI
Press.
Quifiones, M. A., Ford, J. K., & Teachout, M. S. (1995). The relationshipbetween work experience and job
performance:a conceptualand meta-analyticreview. Personnel Psychology, 48, 887-910.
*Rao, A., Schmidt, S. M., & Murray,L. H. (1995). Upwardimpressionmanagement:goals, influence strategies,
and consequences. HumanRelations, 48, 147-167.
Schmidt, F. L., & Hunter,J. E. (1996). Measurementerrorin psychological research:lessons from 26 research
scenarios. Psychological Methods, 1, 199-223.
*Stevens, C. K., & Kristof, A. L. (1995). Making the right impression: a field study of applicant impression
managementduringjob interviews.Journal of Applied Psychology, 80, 587-606.
Tedeschi, J. T. (Ed.). (1981). Impression management theory and social psychological research. New York:
Academic Press.
Tedeschi, J. T., & Melburg,V. (1984). Impressionmanagementin the organization.Research in the Sociology of
Organizations,3, 31-58.
*Thacker,R. A. (1995). Gender, influence tactics, and job characteristicspreferences:new insights into salary
determination.Sex Roles, 32, 617-638.
*Thacker,R. A. (1999). Perceptionsof trust,upwardinfluence tactics, and performanceratings.Perceptualand
Motor Skills, 88, 1059-1070.
*Thacker,R. A., & Wayne,S. J. (1995). An examinationof the relationshipbetween upwardinfluencetactics and
assessments of promotability.Journal of Management,21, 739-756.
*Valerius, L. (1990). Antecedents of impression management among managers in public leisure service
organizations and the effects of impression management on performanceappraisal (Doctoral dissertation,
Universityof Illinois, 1990). Dissertation AbstractsInternational,52, 226.
Villanova, P., & Bernardin,H. J. (1991). Performanceappraisal:the means, motive, and opportunityto manage
impressions.In R. A. Giacalone, & P. Rosenfeld (Eds.), Applied impressionmanagement:How image-making
affects managerial decisions (pp. 81-96). Newbury Park,CA: Sage Publications.
Viswesvaran,C., Ones, D. S., & Schmidt,F L. (1996). Comparativeanalysis of the reliabilityof job performance
ratings.Journal of Applied Psychology, 81, 557-574.
*Watt,J. D. (1993). The impactof the frequencyof ingratiationon the performanceevaluationof bankpersonnel.
The Journal of Psychology, 127, 171-177.
*Wayne, S. J., & Ferris,G. R. (1990). Influence tactics, affect, and exchange quality in supervisor-subordinate
relations:a laboratoryexperimentand a field study.Journal of Applied Psychology, 75, 487-499.
*Wayne, S. J., & Kacmar,K. M. (1991). The effects of impression managementon the performanceappraisal
process. OrganizationalBehavior and HumanDecision Processes, 48, 70-88.
*Wayne,S. J., & Liden, R. C. (1995). Effects of impressionmanagementon performanceratings:a longitudinal
study.Academyof ManagementJournal, 38, 232-260.
*Wayne, S. J., Liden, R. C., Graf, I. K., & Ferris, G. R. (1997). The role of upwardinfluence tactics in human
resourcedecisions. Personnel Psychology, 50, 979-1006.
Wortman,C. B., & Linsenmeier,J. A. (1977). Interpersonalattractionand ingratiationin organizationalsettings.
In B. M. Staw, & G. R. Salancik(Eds.), New directions in organizationalbehavior(pp. 133-178). Chicago:St.
Clair Press.
the impact of impressionmanagement
*Yates, V. L. (1991). An investigationof careerperformance/progression:
(Doctoral dissertation,Texas A&M University, 1991). DissertationAbstractsInternational,53, 222-223.
Yukl, G., & Falbe, C. M. (1990). Influence tactics and objectives in upward,downward,and lateral influence
attempts.Journal of Applied Psychology, 75, 132-140.
Yukl,G., & Falbe,C. M. (1996). Antecedentsof influenceoutcomes.Journalof AppliedPsychology,81, 309-317.
*Yukl,G., & Tracey,J. B. (1992). Consequencesof influencetactics used with subordinates,peers, and the boss.
Journal of Applied Psychology, 77, 525-535.
Copyright ) 2003 John Wiley & Sons, Ltd.
J. Organiz.Behav. 24, 89-106 (2003)