Robert Hogan, Gordon J. Curphy, and Joyce Hogan

Although psychologists know a great deal about leader-ship, persons who make decisions about real leaders seemlargely to ignore their accumulated wisdom. In an effortto make past research more accessible, interpretable, andrelevant to decision makers, this article defines leadershipand then answers nine questions that routinely come upwhen practical decisions are made about leadership (e.g.,whom to appoint, how to evaluate them. when to terminatethem).

identify with the welfare of our social unit-perhaps be-cause individual survival sometimes depends on groupsurvival (Eibl-Eibesfeld, 1989; J. Hogan, 1978). It is im-portant to distinguish between a person's short-term andlong-term self-interest; actions that promote the groupalso serve an individual's long-term welfare. Historymournfully suggests, however, that without an externalthreat to their group, people largely pursue their short-term interests.

This article provides a context for understandingleadershi!p. In our view, leadership involves persuadingother peclple to set aside for a period of time their indi-vidual cc)ncerns and to pursue a common goal that isimportaDlt for the responsibilities and welfare of a group.This defi)ution is morally neutral. A Somali warlord whois trying 1:0 bring together a group of clansmen to controlfood suPJ;>lies needs the same skills as an inner-city Chi-cago minister who is trying to bring together a group ofparishionlers to help the homeless.

Leadership is persuasion, not domination; personswho can require others to do their bidding because oftheir power are not leaders. Leadership only occurs whenothers willingly adopt, for a period of time, the goals ofa group as their own. Thus, leadership concerns buildingcohesive and goal-oriented teams; there is a causal anddefinitional link between leadership and team perfor-mance.

What is it that leaders do? Beginning with the OhioState stu(iies in the 1940s and I 950s, several taxonomiesof leadership behaviors have been proposed, includingthose by Borman and Brush (1993), Davis, Skube, Hel-lervik, G,~belein, and Sheard (1992), and Yuki, Wall, andLepsinger (1990). They differ primarily in terms of theirspecificity. Yuki et al.'s list is the broadest; it identifies 14categories of leader behavior, including planning and or-ganizing, problem solving, clarifying, informing, moni-toring, motivating, consulting, recognizing, supporting,managing conflict and team building, networking. dele-gating, dlevelOPing and mentoring, and rewarding. Al-

Robert Hogan and Joyce Hogan, Department of Psychology, University ofTulsa. Gorclon J. Curphy, Perronnel Decisions, Inc.. Minneapolis, MN.

David P. Campbell served as action editor for this article.The a'ilthors would like to thank David Campbell, John Campbell,

Dianne Nilsen, Mark Schmit, Robert Smither, and five anonymous re-viewers for their thoughtful comments on earlier drafts of this article.

Correspondence concerning this article should be addressed toRobert Hogan. Department of Psychology, University of Tulsa, Tulsa,OK 74104.

A ccording to the political scientists, the funda-

mental question in human affairs is "Who shall

rule?" As psychologists-who are less infusedwith the spirit of realpolitik-we believe the question is"Who should rule?" The question must be answered dur-ing national elections, when CEOs are replaced, and whenuniversity presidents retire. The question concerns howto evaluate leadership potential. When it is answered in-correctly, teams lose, armies are defeated, economiesdwindle, and nations fail.

In terms of the number of printed pages devoted tothe subject, leadership appears to be one of the most im-portant issues in applied psychology. Volumes appear onthe topic every year, and a recent review lists over 7,000books, articles, or presentations (Bass, 1990). However,the rules of psychological research are such that we tendto focus on narrowly defined issues. The result is that ourresearch is primarily read by other psychologists. Al-though J. P. Campbell (1977) and Mintzberg (1982) rec-ommended that researchers pay more attention to appli-cations, what we know seems to have little impact on thepeople who actually make decisions about leadership. Thegap between what we know and what leadership decisionmakers want to know may explain the popularity of suchbooks as In Search oj Excellence (Peters & Waterman,1982), The Change Masters (Kanter, 1983), Leaders: TheStrategies Jar Taking Charge (Bennis & Nanus, 1985),and The New Leaders (A. M. Morrison, 1992). Thesebooks are not intended to be scientific dissertations;rather, they offer practical suggestions about how to iden-tify and evaluate leadership. To reduce the gap betweenresearchers and the lay public, this article answers ninequestions that psychologists are often asked by personswho must choose or evaluate leaders.

What Is Leadership?Various writers have argued that our evolutionary historymakes us both selfish (Dawkins, 1976) and yet able to

493June 1994 .American PsychologistC~ght 1994 by the American Psychological Association. Inc. 0003-066X/94/S2.00Vol. 49. No.6. 493-504

though these actions are required by persons ranging fromfirst-line supervisors to CEOs, their relative importancediffers by organizational level.

These taxonomies tell us what people in leadershippositions typically do, and the various commerciallyavailable, multirater assessment instruments (e.g., Per-sonnel Decisions, Inc., 1991) tell us about the degree towhich a particular leader does these things. However, thereis little published research concerning what effectiveleaders actually do. Effectiveness concerns judgmentsabout a leader's impact on an organization's bottom line(i.e., the profitability of a business unit, the quality ofservices rendered, market share gained, or the win-lossrecord of a team). Indices of effectiveness are often hardto specify and frequently affected by factors beyond aleader's control. Nevertheless, effectiveness is the standardby which leaders should be judged; focusing on typicalbehaviors and ignoring effectiveness is an overarchingproblem in leadership research.

Does Leadership Matter?

In 1910, the Norwegians and the English engaged in adramatic and highly publicized race to the South Pole. Itwas an epic contest, and the contrast between the per-formance of the Norwegian team led by Roald Amundsenand the English team led by Robert Falcon Scott pro\ideda real-life study in leadership and team performance.Scott's incompetence cost him the race, his life, and thelives of three team members, although, as often happenswhen high-level leadership fails, the details were coveredup for years (cf. Dixon, 1976).

The fact that Lincoln's army was inert until UlyssesS. Grant assumed command and that some coaches canmove from team to team transforming losers into winnersis, for most people, evidence that leadership matters. Psy-chologists, as researchers, are (properly) more skeptical;they often explain differences in effectiveness in terms ofthe factors in the "environment" in which a team oper-ates. Perhaps because effectiveness is influenced by somany factors, there are only a handful of studies evalu-ating the impact of leadership on an organization's bottomline. Some of the best e\idence we have concerns the per-formance of flight crews (Chidester, Helmreich, Grego-rich, & Geis, 1991), military units (Curphy, 1991, 1993),U. S. presidents (House, Spangler, & Woycke, 1991), andMethodist ministers (Smith, Carson, & Alexander, 1984).These studies show that certain leader characteristics areassociated with enhanced team performance-when theappropriate indices of effectiveness are studied.

There is a second and less direct way of answeringthe question "Does leadership matter?" At the historicallevel one might reflect on the horrific consequences ofthe leadership of Adolph Hitler in Germany from 1933to 1945 and Joseph Stalin in Russia from 1927 to 1953.Millions of people suffered and died as a consequence ofthe megalomaniacal visions of these two flawed geniuses,and the baleful consequences of their rule persist even

today.

A third way to decide whether leadership matters isto as!" the consumers of leadership (i.e., a manager's directrepoIts). Several patterns of leadership behavior are as-sociated with subordinates' performance and satisfaction(cf. B,ass, 1990; Hughes, Ginnett, & Curphy, 1993; YukI,1989). Conversely, reactions to inept leadership includeturnover, insubordination, industrial sabotage, and mal-ingering. R. Hogan, Raskin, and Fazzini (1990) notedthat ,organizational climate studies from the mid-1950sto the present routinely show that 60% to 75% of theemployees in any organization-no matter when or wherethe survey was completed and no matter what occupa-tional group was involved-report that the worst or moststres~;ful aspect of their job is their immediate supervisor.Goo<lleaders may put pressure on their people, but abu-sive and incompetent management create billions of dol-lars ()flost productivity each year. Dixon's (1976) book,The Psychology of Military Incompetence, provides agraphic and almost unbearably painful account of theconSt~quences of bad leadership in the military. Reactionsto iru~pt leadership can be extreme. In the spring of 1993articles in several major newspapers (e.g., the New YorkTimE~s, the Washington Post) noted that poor first-linesupel~sion was associated with the deaths of numerouspos~LI workers over the past decade.

To stimulate research on the topic of inept manage-ment, R. .Hogan et al. (1990) proposed that the base ratefor managerial incompetence in America is between 60%and 75%. DeVries (1992), in a fascinating brief review,estiIIlated that for the past 10 years the failure rate amongsenic.r executives in corporate America has been at least50%. Shipper and Wilson (1991), using data from 101departments in a large southwestern hospital, reportedthat the base rate for incompetent management in thatorgaIlization was 60%. Millikin-Davies (1992), using datafrom a large aerospace organization, estimated a 50% baserate. She gathered critical incidents of managerial incom-petence, which she rank ordered in terms of frequency.The most common complaints from direct reports con-cernl~ (a) managers' unwillingness to exercise authority(e.g., "is reluctant to confront problems and conflict";"is not as self-confident as others"), which characterized20% of the sample of 84 managers, and (b) managerstyrannizing their subordinates (e.g., "manages his/heremployees too closely, breathes down their necks"; "treatsemployees as if they were stupid"), which characterized16% of the sample.

In summary, a growing body of evidence supportsthe (~ommon sense belief that leadership matters. Con-sequently, psychologists need to better determine when,wheJ:e, and how ~eadership affects organization effective-ness and help organizations choose better leaders.

How Are Leaders Chosen?Psychologists have known for some time that measuresof cognitive ability and normal personality, structuredinteJ:views, simulations, and assessment centers predictleadl~hip success reasonably well (cr. Bass, 1990; Howard& Bray, 1990; Hughes et al., 1993; Sorcher, 1985; YukI,

June 1994 .American Psychologist494

to evaluate leadership effectiveness. Sweetland's (1978)review of managerial productivity concluded that effectiveleadership and increased group output were a functionof the interaction between managers and their subordi-nates. MtLrphy and Cleveland (1991) noted that the eval-uation of a manager's performance depends, in part, onthe relationships that the person has established with hisor her subordinates. Hegarty (1974) found that universitydepartment chairs who received feedback from subor-dinates iraproved their performance, both as judged bysubordinates and in comparison with control participantswho received no subordinate evaluations. Similarly, Ber-nardin arId Klatt (1985) found that managers who wereinvolved in multirater appraisal systems received signif-icantly higher mean effectiveness ratings than those whoreceived ][10 subordinate feedback. McEvoy and Beatty(1989) compared the predictive validity of subordinateevaluations with assessment center ratings and concludedthat subordinate ratings were as effective (and less expen-sive) as ~;sessment center data in forecasting managerialperformance seven years later.

Becaluse subordinates are in a unique position tojudge leadership effectiveness, what leadership character-istics do they feel are most important? Research by D. P.Campbelll (1991), Harris and Hogan (1992), and Lom-bardo, RILlderman, and McCauley (1988) indicates thata leader's credibility or trustworthiness may be the singlemost impfOrtant factor in subordinates' judgments of hisor her effe:ctiveness. For example, Harris and Hogan askedsubordinates (N = 30 I) to evaluate their managers (N= 49) usi:t1g a 55-item questionnaire that assessed growth

versus stagnation, interpersonal competence, managerialvalues, and technical competence. Subordinates also ratedtheir managers for overall effectiveness. Each managerand his o,r her boss completed a parallel questionnaire.Subordinates' and bosses' evaluations of a target man-ager's performance were reasonably consistent (~ > .50).In addition, managers' self-ratings were uncorrelated withthe rating$ provided by the other groups; this is consistentwith the meta-analytic results of Harris and Schaubroeck(1988). Perhaps most important, bosses' ratings of amanager's overall effectiveness were largely influenced byjudgmen1:S of his or her technical competence (e.g., "Su-pervisor is a flexible and far-sighted problem solver"),whereas ~.ubordinates' ratings of a manager's overall ef-fectivene~;s were largely influenced by judgments of in-tegrity (e.g., "My supervisor has earned my trust"). Thus,although subordinates and bosses tend to agree in theirevaluation of a manager's overall effectiveness, they alsoevaluate rather different aspects of that performance. Al-though slJbordinates' ratings will be to some degree con-taminated by rating errors, research shows that these rat-ings also reflect some knowledge of a person's actualperform~lllce in a leadership role. For example, Shipperand Wilson (1991), using data provided by managers andtheir sutlordinates from 68 subunits of a large south-western hospital, showed that subordinates' ratings ofmanagerial effectiveness were correlated (rs between .22and .46) with engineered standards of productivity. These

495June 1994 .American Psychologist

Hogatl, 1992). Finally, because subordinates', peers', orbosse:;' ratings involve judgments about the frequency ofcertain behaviors, researchers typically find stronger linksbetween personality and these ratings than between per-sonality and indices of effectiveness.

Wh,f Do We Choose So Many FlawedLea.~ers?The 1992 U. S. presidential election is an example of howimpol.1ant leaders are often chosen. A group of candidatesmake public statements; the voters, aided by promptingsfrom .journalists, evaluate the leadership potential of thecandi,:iates and then chose one. The process involves es-timarilng the leadership qualities of strangers. De Vries(1992) noted that, of all the methods available to chosesenior executives, organizations overwhelmingly rely onsearch firm nominations, background checks, and inter-views. The standardized and well-validated methods de-velOpt~d by psychologists are used in only a tiny fractionof cases. We believe the less valid methods continue tobe use:d (in spite of what we know) because of the reasonscited earlier and because candidates for executive posi-tions often refuse to submit to psychological assessment.The 50% failure rate among senior executives may wellbe th(~ result of these widely used but invalid selectionproce.:iures. Again, the hiring problem typically involvesevaluating"the leadership potential of strangers.

l:bere has been considerable research concerning thecharal::teristics of persons who, in the absence of perfor-manCt~ data, nonetheless seem leaderlike. This researchfits nicely into two categories. These include (a) studiesof the relation between personality and indices of emer-gent 1~dership, and (b) research on implicit leadershiptheo~l.

Emergent Leadership

Research on emergent leadership identifies the factors as-sociated with someone being perceived as leaderlike whenthere is only limited information about that person's ac-tual ~rformance; this research is typically involves lead-erless ,discussion groups. Stogdill (1948) reviewed researchon personality and emergent leadership in a variety ofunstnlctured groups. He concluded that measures ofdominance, extraversion, sociability, ambition orachievement, responsibility, integrity, self-confidence,mood and emotional control, diplomacy, and coopera-tiveness were positively related to emergent leadership.

l~he personality descriptors identified in Stogdill's(1974) review easily map onto the big-five model of per-sonality structure endorsed by many modern personalitypsychologists (cf. Digman, 1990; Goldberg, 1993; RHo-gan &: Hogan, 1992; McCrae & Costa, 1987; Passini &Norman, 1966). This model holds that personality, fromthe vil~w of an observer, can be described in terms of fivebroad dimensions-surgency, agreeableness, conscien-tiOUSI:less, emotional stability, and intellect-and it pro-vides a common vocabulary for interpreting the resultsof per:)Onality research. In the past, this research was often

findings provide strong support for the use of subordi-nates' evaluations of managerial effectiveness.

A third category of studies evaluates the leadershippotential of strangers on the basis of their performancein interviews, simulations, assessment centers, or lead-erless group discussions. Examples include studies by AI.bright, Kenny, and Malloy (1988), Howard and Bray(1990), and Lord, De Vader, and Allinger (1986). Theleaderless group research provides virtually no infor-mation about effectiveness; rather, it tells us about whata person must do in order to be perceived, in the shortterm, as leaderlike. On the other hand, assessment centerresearch often uses organizational advancement as a cri-terion, and it tells us about the characteristics related togetting ahead in large, complex organizations. In theAT&T Managerial Assessment Project, for example, sub-sequent management level was best predicted by assess-ment center ratings for need for advancement, generalmental ability, written communication skills, overallcommunications skills, flexibility, creativity, and orga-nizing and planning (Howard & Bray, 1990).

Fourth, self-ratings of leadership have also been usedas evaluative criteria (Farh & Dobbins, 1989). The evi-dence is clear, however, that self-ratings tell us little aboutleader effectiveness. But there is a kind of manager whoroutinely over evaluates his or her performance, and thattendency is associated with poor leadership (Atwater &Yammarino, 1992; Nilsen & Campbell, 1993; Van Velsor,Taylor, & Leslie, 1992).

In the fifth category of research, effectiveness is de-fined by the low end of the continuum-by persons whosecareers are in jeopardy or who have derailed. The factthat a person has been passed over for promotion or firedreflects an evaluation of his or her performance in a neg-ative direction. Early research on derailment includes ar-ticles by Lombardo et al. (1988) and McCall and Lom-bardo (1983). Hellervik, Hazucha, and Schneider (1992),Peterson (1993), and Peterson and Hicks (1993) studiedmanagers whose careers were in trouble, using a widevariety of assessment techniques, such as multirater as-sessment instruments and psychological tests, to identifydifferent jeopardy and derailment factors. This researchreveals managerial incompetence to be associated withuntrustworthiness, over control, exploitation, micro-management, irritability, unwillingness to use discipline,and an inability to make good staffing or business deci-sions (or both). .

We can summarize this section as follows: The an.swer to the question "How should leaders be evaluated?"is "In terms of the performance of their teams." Realis-tically, the data needed to make this evaluation are oftendifficult to obtain or badly contaminated by external fac-tors. Perhaps the best alternative is to ask subordinates,peers, and superiors to evaluate a leader. The empiricalliterature suggests that these sources of information arecorrelated; that the respondents tend to key on differentaspects of a leader's performance; and that, taken together,these evaluations are moderately but significantly relatedto team performance (D. P. Campbell, 1991; Harris &

496 June 1994 .American Psychologist

hard to interpret because different studies used differentterminology. For example, the conscientiousness dimen-sion has been called conformity (Fiske, 1949), prudence(R. Hogan & Hogan, 1992), constraint (Tellegen, 1982),will to achieve (Digman, 1990), and work (Peabody &Goldberg, 1989). These different terms refer to the samebroad construct; similar trends for the other four person-ality dimensions can be found in the Appendix.

Returning to Stogdill's (1948) review, dominance,e~traversion, and sociability reflect surgency; responsi-bility, achievement, and integrity fall into the conscien-tiousness dimension; self-confidence, mood, and emo-tional control are part of emotional stability; and diplo-macy and cooperativeness resemble agreeableness. Mann(1959) reviewed 28 studies concerning the relation be-tween personality and observer ratings of emergent lead-ership in small groups and essentially replicated Stogdill's(1948) findings.

More recent studies of personality and leadershipemergence reached similar conclusions (Gough, 1990; J.Hogan, 1978; Kenny & Zaccaro, 1983; Lord et al., 1986;Rueb & Foti, 1990; Stogdill~ 1974; Zaccaro, Foti, &Kenny, 1991). Gough (1984, 1990), for example, reportedthat the Dominance, Capacity for Status, Sociability, So-cial Presence (i.e., surgency), Self-Acceptance, Achieve-ment via Independence (i.e., emotional stability), andEmpathy (i.e., agreeableness) scales of the California Psy-chological Inventory (Gough, 1987) are significantly cor-related with peer and staff ratings for emergent leadershipin leaderless discussion groups. Lord et al. (1986) usedmeta-analysis to estimate the correlations between variouspersonality traits and leadership emergence in the studiesreviewed by Mann (1959) and 13 other studies. They re-ported that the "true" correlations between masculinity-femininity, dominance, extraversion-introversion, ad-justment, conservatism (which correspond to surgency,emotional stability, and conscientiousness), and leader-ship emergence in small groups are r = .34, .13, .26, .24,and .22, respectively. Although the correlations tend tobe low, many studies in this meta-analysis examinedleadership emergence in a single situation, and these singlesituation ratings are necessarily less reliable than ratingsfrom a variety of situations.

Looking across a number of leaderless discussiongroups, Kenny and Zaccaro (1983) found that between48% and 82% of the variance in leadership emergencerankings was due to personality. Ellis (1988), Rueb andFoti (1990), and Zaccaro et al. (1991) have shown thatthe ability to control one's expressive behaviors (i.e., self.monitoring) is positively related to leadership emergence.Snyder's (1974) self-monitoring scale consists of three di-mensions-concern for social appropriateness, sensitivityto social cues, and the ability to control one's behavioraccording to social cues (cf. Briggs, Cheek, & Buss,1980)-and these dimensions correspond to the big-fivedimensions of surgency, agreeableness, and emotionalstability. In summary, this research reveals a fairly con-sistent association between high scores on the dimensionsof surgency, agreeableness, and emotional stability and

being perceived as leaderlike in a group with no appointedleader.

As..'iessment centers and employment interviews areoften ~:d to evaluate the leadership potential of strangers;data are sometimes gathered to determine the validity ofthese ev~l1uations. Bray (1982), for example, reported thatassessmt:nt center data were reasonably valid predictorsof a person's promotion record at AT&T. The perfor-mance climensions identified in the AT&T ManagerialAssessment Project-need for advancement, behaviorflexibility, creativity, organizing and planning, and soon-correspondto the dimensions of surgency, conscien-tiousnes:;, emotional stability, and intellect.

Tht~ results suggest that the big-five model providesa conveJuent way to summarize both leaderless groupdiscussic.n and assessment center research. The resultsalso suggest that measures of surgency, agreeableness,conscientiousness, and emotional stability can be used topredict the leadership potential of strangers. Implicitleadershiip theory research also supports the utility of thebig-five taxonomy.

Implicit leadership Theory

The seccmd line of research concerning how we evaluatethe leadership potential of strangers is called implicitleadership theory. Starting with Hollander and Julian(1969), implicit leadership theory argues that people areseen as leaderlike to the degree that their characteristics(i.e., intelligence, personality, or values) match other peo-ples' pret:onceived notions of what leaders should be like.Eden and Leviathan (1975), Lord, Foti, and De Vader(1984), Rush, Thomas, and Lord (1977), and Weiss andAdler (1 ~)81) have shown that people do in fact have gen-eralized ideas about leadership, and they use them toevaluate the leadership potential of strangers. Specifically,most pe<>ple seem to regard intelligence, honesty, socia-bility, understanding, aggressiveness, verbal skills, deter-mination, and industriousness as important aspects ofleadership, regardless of the team task or situation. Notethat these attributes can be organized using the big-fivemodel.

To return, finally, to the question of this section,"Why do we choose so many flawed leaders?," the answermay be that search committees choose candidates not onthe basis of established principles of personnel selectionbut on the basis of the principles that guide leadershipemergence-namely, those candidates who seem mostleaderlike are most likely to be anointed. The problem isthat pen:ons who seem leaderlike may not have the skillsrequired to build and guide an effective team. The resultis a leadership failure rate in the range of 50% to 60%.

How to Forecast ,Leadership?

In our jlLdgment, the best way to forecast leadership is touse a combination of cognitive ability, personality, sim-ulation, role play, and multirater assessment instrumentsand tech:niques. Although personality assessment is partof this, there is some disagreement as to whether person-ality measures on their own can predict leadership po-

497June 1994 .American Psychologist

memt>ers, such leaders are able to persuade them to workfor and to support the vision.

Charismatic leaders can be quite effective; relativeto noncharismatic leaders, they have substantially higher(a) promotion recommendations or performance ap-praisal ratings from superiors; (b) satisfaction, morale, orappro'val ratings from subordinates; (c) historians' ratingsof greatness; or (d) levels of team performance (Avolio,Waldman, & Einstein, 1988; Bass, 1985; Bass, Avolio, &Goodlheim, 1987; Bass & Yammarino, 1991; Conger &KanuJrlgo, 1988; Curphy, 1991, 1993; House et al., 1991;Howe:ll & Frost, 1988).

House et al. (1991) reported that charismatic U.S.presidents have strong needs for power and high energylevels and they are socially assertive (these themes resem-ble surgency) and achievement oriented (i.e., conscien-tiousness). Using self-ratings from the Adjective Checklist(Gougj1 & Heilbrun, 1983) and subordinates' ratings forcharisma, Ross and Offermann (1991) reported thatcharisma ratings are positively correlated with self-con-fideno~ and personal adjustment (i.e., emotional Stability),femin:lne attributes and nurturance (i.e., agreeableness),and trle need for change (i.e., intellect).

Foushee, Chidester, Helmreich, and their associatesstudied the personality measures that influence team per-formance-in this case, the performance of commercialairline: flight crews (cf. Chidester et al., 1991; Foushee &Helmreich, 1988). This research is important becausebreakdowns in team performance are the primary causeof air transport accidents (Cooper, White, & Lauber,1979). Chidester et al. showed that flight crew perfor-man~~-defined in terms of the number and severity ofthe enrors made by the crew-is significantly correlatedwith me personality of the captain. Crews with captainswho \N'ere warm, friendly, self-confident, and able to standup to pressure (i.e., agreeableness and emotional Stability)made the fewest errors. Conversely, crews with captainswho ~'ere arrogant, hostile, boastful, egotistical, passiveaggressive, or dictatorial made the most errors. Despitethese results, Chidester et al. pointed out that personalityis not taken into account in the process of airline pilotselection.

Wh)r Do Leaders Fail?

Leaders fail for a variety of reasons-product lines nolonger interest customers, services are no longer required,and companies reorganize and downsize. Nevertheless, anumber of leaders fail for personal rather than structuralor economic reasons. They may be skilled in a particulararea, !;uch as accounting, engineering, or sales. They failbecause they can no longer rely solely on their own skillsand effort; that is, they have been promoted into positionsthat re:quire them to work through others to be successful.Because they are unable to build a team, their manage-ment careers come to a halt. Derailment is curiously un-derstudied given the frequency with which it occurs.

Bentz (1985) essentially founded modern derailmentresearch while an~yzing the correlates of executive per-forma.nce at Sears. He reported that among the persons

tential. We believe that terminological confusions haveobscured the usefulness of personality measures for as-sessing leadership potential and that the big-five modelsubstantially enhances our ability to integrate this re-search.

Personality and Rated Leader Effectiveness

Several lines of evidence show that certain personalitydimensions are consistently related to rated leadershipeffectiveness. The first evidence comes from Stogdill's(1974) review. Stogdill found that surgency (i.e., domi-nance, assertiveness, energy or activity level; speechfluency, sociability, and social participation), emotionalstability (i.e., adjustment, emotional balance, indepen-dence, and self-confidence), conscientiousness (i.e., re-sponsibility, achievement, initiative, personal integrity,and ethical conduct), and agreeableness (i.e., friendliness,social nearness, and support) were positively related torated effectiveness. Stogdill (1974) did not organize hisfindings as we describe them; nonetheless, his findingssupport the idea that there is a relationship between per-sonality and leadership.

Bentz (1985, 1987, 1990) reported similar findingsfrom his research on executive selection at Sears. Usingthe Guilford-Martin Personality Inventory, Bentz (1985,1990) noted that executives promoted to the highest levelsat Sears were articulate and active (i.e., surgency), inde-pendent, self-confident, and emotionally balanced (i.e.,emotional stability), and hard working and responsible(i.e., conscientiousness). The median multiple correla-tions between these dimensions and subordinates' ratingsof operating efficiency, personal relations, satisfaction, fi-nancial rewards, and job conditions were about R = .50(Bentz, 1985). Bentz (1985) reported comparable multiplecorrelations between these personality factors and leaders'compensation, immediate and second-level superiors'ratings and rankings, and peer groups' ratings of effec-tiveness over a 2 I-year period.

Bray and Howard (1983) and associates reportedsimilar findings with AT&T executives. Those personalitytraits that best predicted managerial advancement-andwe assume that most of those who advanced were alsoeffective-were the desire for advancement, energy-ac-tivity level, and the readiness to make decisions (i.e., sur-gency); resistance to stress and tolerance for uncertainty(i.e., emotional stability); inner work standards (i.e., con-scientiousness); and range of interests (i.e., intellect; Bray,Campbell, & Grant, 1974; Bray & Howard, 1983; Howard& Bray, 1990).

Personality and Effective Team Performance

Two sets of studies illustrate the link between personalityand team performance. The first concerns charismaticleadership. House (1977) used biographical materials toidentify three themes in the careers of charismatic leaders.First, they have a vision that others find compelling; sec-ond, they are able to recruit a group of people who sharethat vision, and these people resemble a team; and third,by virtue of the relationships they develop with the team

June 1994 .American Psychologist498

mance; six problems are task related and two involveteam maintenance. On the task side, successful leaderscommunicate a clear mission or sense of purpose, identifyavailable resources and talent, develop the talent, planand organize, coordinate work activities, and acquireneeded rl~sources. On the maintenance side, they mini-mize and! resolve conflicts among group members andthey ensure that team members understand the team'sgoals, constraints, resources, and problems. These team-building tasks obviously overlap with the taxonomy ofleader behaviors developed by Yuki et al. (1990).

We believe that a leader's personality has predictableeffects on team performance. For example, leaders withhigher surgency scores communicate more with theirteams, which increases the possibility that the team un-derstands its goal and the performance standards requiredto achieve it. Moreover, these leaders are better able tobuild alliances with people outside of the team, whichallows the:m to secure necessary equipment and resources.Conscientiousness is related to being perceived as trust-worthy, planful, and organized. Agreeableness is relatedto communication, trust, and morale. Emotional stabilityis associated with seeming steady under pressure, able toresolve ccmflicts, and to handle negative feedback, all ofwhich promote team effectiveness. Thus, four of the fivebig-five personality dimensions are related to Hallam andCampbell's (1992) team building tasks and to YukI's(1989) taxonomy of leader behaviors.

The discussion so far concerns the relationship be-tween personality and leadership in general. But practi-tioners also know that leadership is relative to the groupin question. Although there are few data available on thispoint, we suspect that two considerations are important.The first is the group's developmental history; the secondconcerns the major tasks the group performs.

One can speculate that the qualities needed to forma group DllaY be different from those required to maintainit. Persons leading organizations in the start-up phase maybe more effective if they have a credible and strategicvision of what the group can do. Such people will alsoneed to withstand the discouragement associated withthe inevi1table failures in the start-up phase. On the otherhand, su(;cessfulleaders of established organizations willprobably need to be more orderly, more ceremonial, moreconcerned about details, and less visionary. Thus, leaderswith higher surgency, intellect, and emotional stabilityscores may be more successful in organizations developingnew products or services, whereas leaders with higherconscientiousness scores may be more effective in orga-nization!i having established products, services, and pro-cedures.

Leadership is also relative to the task of the group,but how should that relationship be conceptualized? Hol-land's (11~85) theory of occupational types provides a clue.Holland proposed that people's interests, talents, abilities,values, a.nd motives cluster in six broad types. Realistictypes (e.!~., engineers) are procedural, action oriented, andconcrete. Investigative types (e.g., scientists) are abstract,original, and independent. Artistic types (e.g., painters,

with the appropriate positive characteristics (i.e.; intelli-gence, confidence, ambition), a subset failed. Bentz cat-alogued the themes associated with failure (e.g., playingpolitics, moodiness, dishonesty) and concluded that thefailed executives had an overriding personality defect orcharacter flaw that alienated their subordinates and pre-vented them from building a team.

Research on managerial incompetence at the Centerfor Creative Leadership and Personnel Decisions, Inc.,has come to similar conclusions; many managers who arebright, hard-working, ambitious, and technically com-petent fail (or are in danger of failing) because they areperceived as arrogant, vindictive, untrustworthy, selfish,emotional, compulsive, overcontrolling, insensitive,abrasive, aloof, too ambitious, or unable to delegate ormake decisions (Hazucha, 1991, Kaplan, Drath, & Ko-fodimos, 1991; Lombardo et al., 1988; McCall & Lom-bardo, 1983; Peterson & Hicks, 1993).

The big-five model reflects the "bright side" of per-sonality. Effectiveness requires both the presence of thesepositive characteristics and the absence of what we call"dark side" characteristics-irritating tendencies that al-ienate subordinates and interfere with a person's abilityto form a team. Research shows that these dark side char-acteristics are negatively related to ratings of team per-formance and that subordinates are almost always awareof them (Harris & Hogan, 1992). Nonetheless, they arehard to detect using interviews, assessment centers, orinventories of normal personality because they coexistwith high levels of self-esteem and good social skills (Har-ris & Hogan, 1992). Because managers with dark sidetendencies often do well in procedures that evaluate theleadership potential of strangers, their counterproductivetendencies will be apparent only after they have been onthe job for some time.

Can dark side characteristics be changed? The bestevidence here comes from an evaluation of the IndividualCoaching for Effectiveness program at Personnel Deci-sions, Inc.-an intensive intervention that may last for ayear. The program is designed for managers whose careersare in jeopardy. Reviewing the results for 370 candidatesover a five-year period, Peterson (1993) and Peterson andHicks (1993) reported that the majority of managers wereable to change a number of targeted behaviors, and thesebehavioral changes were still in place six months afterthe training had ended. Many of these candidates hadpreviously attended three- to five-day standardized lead-ership training programs, but these programs producedlittle behavioral change (Peterson, personal communi-cation, November 18, 1993). These findings imply thatmany managers who are performing poorly can make thechanges necessary to maintain their careers, but they needmore intensive training than that found in most leadershipdevelopment programs.How Do leaders Build Teams?The key to a leader's effectiveness is his or her ability tobuild a team. Hallam and Campbell (1992) identifiedeight problems for leadership that affect team perfor-

compt~tition for talented employees will increase and be-cause managerial responsibilities will expand, the overallqualit:r of management will need to improve. Corporatefailures are increasing-there were 57,000 failures in 1986(Ropp, 1987)-and this may reflect the combination ofincompetent management and changes in labor and mar-ket forces. If current estimates of the base rates of badmanagement are realistic, then organizations in which60% or the managers are incompetent will likely be at aseriou~; competitive disadvantage. Psychologists can helporganizations by verifying estimates of the base rate ofincompetent management, exploring the relationshipsbetween these estimates and organizational effectiveness,and by doing a better job informing organizations aboutour m:anagerial selection, coaching, and promotion ex-pertise.

Second, with an increased emphasis on productivity,we suspect that the performance of senior managers willbe mol:e closely scrutinized. Derailed managers are typ-ically E:ood at selling themselves upward in their organi-zation~:, but they are less successful when dealing withpeers and subordinates; .thus these groups have access tounique information. Consequently, if effectiveness be-comes :3. criterion for managerial evaluation, then multipleperspe<;tive appraisals (e.g., those that include bosses',peers', and subordinates' ratings) may become morewidespread. Psychologists have played a key role in thedevelopment and refinement of multirater assessment in-struments, and they should play an equally importantrole in the adoption of these instruments in the future.

Third, management practices will have to change aswe mo've toward a service economy and the workforcebecomt:5 more diverse-what is the best way to managefemale and minority employees in social and artistic (ser-vice) o~ganizations? Moreover, we will likely have the samepercen1ages of women and minorities in management asare cun.ently in the workforce. Are there significant genderor cultural differences in leadership style, and will thesestyles be more or less effective for building teams in to-morrov/'s organizations? These are questions that psy-chologists are uniquely qualified to answer.

Fourth, although psychologiSts know more aboutleadership than the public apparently recognizes or, in-deed, tllan we are often willing to admit (cf. Meindl &Ehrlich, 1987), there is one aspect of leadership aboutwhich we know very little: how to manage creative talent.There is good reason to believe that successful organi-zations will increasingly rely on innovation and the de-velopm'~nt of new products and services-meaning, onthe performance of their investigative and artistic teams.We understand something about the characteristics of in-dividua! creativity (Barron, 1965; Cronbach, 1984), butwe knO\V little about how to manage teams whose primarytasks ale problem solving and the development of newknowle<ige, methods, and products (cf. J. D. Morrison,1993). How to manage creativity is one of the most im-portant problems of the future, and it is a problem towhose solution psychology can make an important con-tribution.

writers, philosophers) are unconventional, nonconform-ing, and imaginative. Social types (e.g., teachers, clergy,personnel managers) are friendly, idealistic; and altruistic.Enterprising types (e.g., lawyers, politicians) are outgoing,assertive, and manipulative. Finally, conventional types(e.g., accountants, computer programmers) are conform-ing, practical, and conservative. Schneider (1987) pro-posed that the culture of an organization depends on theHolland types of the senior management, that people willjoin organizations whose activities and values are consis-tent with their own preferences, and that they leave or-ganizations whose culture is inconsistent with their pref-erences. -

Moving to the level of the work group, R. Hoganand his associates (Driskell, Hogan, & Salas, 1987; R.Hogan, Raza, & Driskell, 1988) showed that teams canbe classified in terms of their primary tasks using theHolland model. Realistic and conventional groups (e.g.,athletic teams, police departments) respond to task-ori-ented and authoritative leadership and resent participa-tory management, which they see as weak. Enterprisingand social groups (e.g., management teams, school fac-ulties) respond to process, interaction, and participationand resent task-oriented leadership, which they see asauthoritarian. In our view, the familiar Fleishman ~dHarris (1962) leadership typology of initiating structureversus consideration only applies to realistic, conven-tional, enterprising, and social groups; we know littleabout the leadership style that is best suited for artisticand investigative groups (e.g., theater companies, researchteams)-meaning, we know little about the process ofmanaging creativity.

What About Leadership in Workforce2000?Historically, the typical American worker has been aWhite man with a high school education employed in amanufacturing (i.e., realistic and conventional) job. Ourmodels of leadership largely focus on how to lead thatkind of person in those kinds of jobs in those kinds oforganizations. All of the projections suggest, however, thatthe economy will shift from manufacturing to service (i.e.,more social and artistic) jobs and that the workforce willbecome older, less well trained, more diverse, and morefemale (Hamilton, 1988; Johnston & Packer, 1987; Of-fermann & Gowing, 1990). For example, "Only 15 per-cent of the new entrants to the labor force over the next13 years will be native white males, compared to 47 per-cent in that category today" (Johnston & Packer, 1987,p. xiii). The labor market for skilled workers will tighten,and there will be increased competition for talented per-sonnel: "The fastest-growing jobs will be in professional,technical, and sales fields requiring the highest educationand skill levels" (Johnston & Packer, 1987, p. xxi). Asorganizations shrink. fewer middle managers will beneeded. and the responsibilities of first-line managers willexpand.

We see these trends as having five implications forleaders, organizations, and psychologists. First, because

500 June 1994 .American Psychologist

Fifth, given that personality measures can predictleadership effectiveness, how can psychologists best usethis information? We recommend selecting personalitypredictors on the basis of job analysis results becausemeasures chosen in this way have significantly higher cor-relations with performance (Tett, Jackson, & Rothstein1991). Next, we recommend matching measures and cri-teria in terms of their specificity (Pulakos, Borman, &Hough, 1988). Although the big-five dimensions are usefulfor summarizing results, they are the wrong band widthfor many prediction problems; narrower measures ofpersonality often yield higher validity coefficients (Cron-bach, 1984; Hough, 1992; Shannon & Weaver, 1949). Wealso recommend screening candidates for dark side ten-dencies using measures of the Diagnostic and StatisticalManual of Mental Disorders, 3rd edirion (DSM-III, Axis2; American Psychiatric Association, 1980) personalitydisorders or using structured interviews with their directreports. Finally, because bad managers often have exag-gerated views of their talents, psychologists may want touse observers' ratings as predictors of leadership potential.Our first two recommendarions often lead to correlarionsin the .20 to .40 range; observers' ratings lead to corre-lations in the .30 to .60 range (Curphy & Osten, 1993;Nilsen, 1992). Although these results are promising, con-siderably more research is needed here.

Finally, as a profession we need to recognize that wecan improve the lives of the incumbents in many orga-nizations (as well as productivity and organizational cli-mate) by improving leadership selection. Nevertheless,organizations will not ask for our help if we continue toargue that there is no such thing as leadership; that lead-ership has little impact on group, team, and organizationaleffectiveness; or that personality and leadership are un-related. Practitioners do not believe these behaviorist-in-spired arguments, and we must get beyond them if wewant to make an impact on important selection decisions.

REFERENCES

Bennis, W. G., & Nanus, B. (1985). Leaders: The Slrategiesfor lakingcharge. New York: Harper & Row.

Bentz, V. J. (1985, A11gust). A view from lhe lOp: A lhirty year perspecliveof research devoled 10 discover;; descriplion. and prediction of exeCUlive

behaviol: Paper presented at the 93rd Annual Convention of theAmerican Psychological Association, Los Angeles.

Bentz, V. J. (1987, A11gust). Contexlual richness as a crilerion consid-eration in personalily research wilh execuzives. Paper presented at the95th Annual Convention of the American Psychological Association,New York.

Bentz, V. J. (1990). Contextual issues in predicting high-level leadershipperformance: Contextual richness as a criterion consideration in per-sonality research with executives. In K. E. Oark & M. B. Oark (Eds.),Measures of leadership (pp. 131-143). West Orange, NJ: LeadershipLibrary of America.

Bernardin, H. J., & Klatt, L. A. (1985). Managerial appraisal systems:Has practice caught up to the state of the art? Personnel AdminislratOI;30, 79-86.

Borgana, E. F. (1964). The structure of personality characteristics. Be-havioral Science, 12, 8-17.

Borman, W. C., & Brush, D. H. (1993). More progress toward a taxonomyofmanageriaI performance requirements. Human Performance. 6. 1-21.

Bray, D, W. (1982). The assessment center and the study of lives. AmericanPsychologisl. 37, 180-189.

Bray, D. W., Campbell, R. J., & Grant, D. L. (1974). Formalive yearsin business: A long-lerm AT&T sludy of managerial lives. New York:

Wiley-Interscience.Bray, D. "f., & Howard, A. (1983). The AT&T longitudinal studies of

managel"S. In K. W. Schaie (Ed.), Longiludinal sludies of adull psy-chologic.czl development (pp. 112-146). New York: Guilford.

Briggs, S. R., Cheek, J. M., & Buss, A. H. (1980). An analysis of theself-moIlitoring scale. Journal of Personalily and Social Psychology,38. 679--686.

Campbell, D. P. (1991). Manual for the Campbell Leadership Index.Minneapolis, MN: National Computer Systems.

Campbell, J. P. (1977). The cutting edge of leadership: An overview. InJ. G. Hlmt & L. L. Larson (Eds.), Leadership: The CUlling edge (pp.221-2~i). Carbondale: Southern Illinois Univmity Press.

Chidester, T. R., Helmreich, R. L., Gregorich, S. E.,& Geis, C. E. (1991).Pilot personality and crew coordination. International Journal ofAviation Psychology, 1. 25-44.

Conger, J. A., & Kanungo, R. N. (1988). Charismatic leadership: Theelusive jaclor in organizational effecliveness. San Francisco: Jossey-Bass.

Cooper, G. E., White, M. D., & Lauber, J. K. (Eds.). (1979). Resourcemanage.rnent on lhe flighl deck (NASA Conference Publication No.2120; NnS No. N80-22083). Moffett Field, CA: NASA-Ames Re-search Center.

Cronbach, L. J. (1984). Essentials of psychologicallesling (4th ed.). SanFrancisco: Harper & Row.

Curphy, G. J. (1991). An empirical invesligation of Bass' (1985) lheoryoftransji7rmalional and lransactional leadership. Unpublished doctoraldissertation, University of Minnesota, Minneapolis.

Curphy, G. J. (1993). An empirical investigation of the effects of trans-formati<>nal and transactional leadership on organizational climate,attrition, and performance. In K. E. Oark, M. B. Oark, & D. P.Campbt'll (Eds.), Impact of leadership (pp. 177-188). Greensboro,NC: Center for Creative Leadership.

Curphy, G. J., & Osten K. D. (1993). Technical manualfor lhe LeadershipDevelopment Survey (Tech. Rep. No. 93-14). Colorado Springs, CO:U. S. Air Force Academy.

Davis, B. ].., Skube, C. J., Hellervik, L. W., Gebelein, S. H., & Sheard,J. L. (1992). Successful manager's handbook: Development suggeslionsfor loday's managers. Minneapolis, MN: Personnel Decisions.

Dawkins, R. (1976). The selfish gene. New York: Oxford University Press.DeVries, D. L. (1992). Executive selection: Advances but no progress.

Issues & Observations, 12. 1-5.Digman. J. M. (1988). Classical theories oflrail organization and lhe

Big Five faclors of personalilY. Paper presented at the 96th AnnualConvention of the American PsychologicaJ Association, Atlanta, GE.

Albright, L., Kenny, D. A., & Malloy, T. E. (1988). Consensus in per-sonality judgments at zero-acquaintance. Journal of Personality andSocial fsychology. 55. 387-395.

American Psychiatric Association. (1980). Diagnostic and statisticalmanual of mental disorders (3rd ed.). Washington, DC: Author.

Atwater, L. E., & Yam marino, F. J. (1992). Does self-other agreementon leadership percej,tions moderate the validity of leadership and per-formance predictions? Personnel Psychology, 45. 141-164.

Avolio, B. J., Waldman, D. A., & Einstein, W. O. (1988). Transformationalleadership in a management game simulation. Group and Organi-zational Studies. 13, 59-80.

Barron, F. (1965). The psychology of creativity. In New directions inpsychology II (pp. 1-134). New York: Holt, Rinehart & Winston.

Bass, B. M. (1985). Leadership and performance beyond expectations.New York: Free Press.

Bass, B. M. (1990). Bass and Stogdill's handbook of leadership (3rded.). New York: Free Press.

Bass, B. M., Avolio, B. J., & Goodheim, L. (1987). Biography and theassessment of transformational leadership at the world class level.Journal of Management. 13. 7-20.

Bass, B. M., & Yam marino, F. J. (1991). Congruence of self and others'leadership ratings of naval officers for understanding successful per-formance. Applied P~h%gy: An International RevieK\ 40. 437-454.

501June 1994 .American Psychologist

Digman, J. M. (1990). Personality structure: Emergence of the five-factormodel. In Annual review of psychology (Vol. 41, pp. 417-440). PaloAlto, CA: Annual Reviews.

Dixon, N. F. (1976). On the psychology of mililary incompetence. London:Futura.

Driskell, J. E., Hogan, R., & Salas, E. (1987). Personality and groupperformance. In C. Hendrick (Ed.), Personalily and social psychologyreview (pp. 91-112). Beverly Hills, CA: Sage.

Eden, I?, & Leviathan, U. (1975). Implicit leadership theory as a de-termmant of the factor Structure underlying supervisory behaviorscales. Journal of Applied Psychology, 60. 736-741.

Eibl-Eibesfeld, I. (1989). Human ethnology, Chicago: Aldine.Ellis, R. J. (1988). Self-monitoring and leadership emergence in groups.

Personality and Social Psychology Bulletin, 14. 681-693.Eysenck, H. J. (1970). The structure of human personality (3rd ed.).

London: Methuen.Farh, J. L., & Dobbins, G. H. (1989). Effects of self-esteem on leniency

bias in self-reports of performance: A structural equation model. Per-sonnel Psychology. 42, 835- 850.

Fiske, D. W. (1949). Consistency of the factorial structUres ofpersonalityratings from different sources. Journal of Abnormal and Social Psy-chology. 44. 329-344.

Fleishman, E. A., & Harris, E. F. (1962). Patterns of leadership behaviorrelated to employee grievances and turnover. Personnel Psychology,15.43-56.

Foushee, H. C., & Helmreich, R. L. (1988). Group interaction and flightcrew performance. In E. L. Weiner & D. C. Nagel (Eds.), Humanfactors in aviation (pp. 189-227). San Diego, CA: Academic Press.

Goldberg, L. R. (1993). The Structure of phenotypic personality traits.American Psychologist. 48. 26-34.

Gough, H. G. (1984). A managerial potential scale for the CaliforniaPsychological Inventory. Journal of Applied Psychology, 69, 233-240.

Gough, H. G. (1987). California Ps}Y:hological Inventory administrator'sguide. Palo Alto, CA: Consulting Psychologists Press.

Gough, H. G. (1990). Testing for leadership with the California Psycho-logical Inventory. In K. E. Oark & M. B. Oark (Eds.), Measures ofleadership (pp. 355- 379). West Orange, NJ: Leadership Library of

America.Gough, H. G., & Heilbrun, A. B., Jr. (1983). The Adjective Checklist

manual: 1983 edilion. Palo Alto, CA: Consulting Psychologists Press.Guilford, J. P. (1975). Factors and factors of personality. Psychological

Bulletin. 82. 802-814.Hallam, G. L., & Campbell, D. P. (1992, May). Selecting team members?

Stan wilh a theory of team effectiveness. Paper presented at the 7thAnnual Meeting of the Society of Industrial and Organizational Psy-chology, Montreal, Quebec, Canada.

Hamilton, M. H. ( 1988, July 10). Employing new tools to recruit workers.Washington Post, pp. HI, H3.

Harris, G., & Hogan, J. (1992, April). Perceptions and personalily cor-relates of managerial effectiveness. Paper presented at the 13th AnnualPsychology in the Department of Defense Symposium, ColoradoSprings, CO.

Harris, M. M., & Schaubroeck, J. (1988). A meta-analysis of self-su-pervisor, self-peer, and peer-supervisor ratings. Personnel Psychology.41,43-62.

Hazucha, J. F. (1991). Success. jeopardy, and performance: Contrastingmanagerial outcomes and their predictors. Unpublished doctoral dis-sertation, University of Minnesota, Minneapolis.

Hegarty, W. H. (1974). Using subordinate ratings to elicit behavioralchanges in supervisors. JoUr!Zal of Applied Psychology. 59, 764-766.

Hellervik, L. W., Hazucha, J. F., & Schneider, R. J. (1992). Behaviorchange: Models, methods, and a review of the evidence. In M. D.Dunnette & L. M. Hough (Eds.), Handbook of industrial and orga-nizational psychology (2nd ed., Vol. 3). Palo Alto, CA: ConsultingPsychologists Press.

Hogan. J. (1978). Personological dynamics of leadership. Journal ofRe-search in Personalitv. 12. 390-395.

Hogan, R., & Hogan,'J. (1992). Hogan Personalily Inventory manual.Tulsa, OK: Hogan Assessment Systems.

Hogan, R., Raskin, R., & Fazzini, D. (1990). The dark side of charisma.In K. E. Oark & M. B. Clark (Eds.), Measures of leadership (pp.343-354). West Orange, NJ: Leadership Library of America.

Hogan, R., Raza, S., & Driskell, J. E. (1988). Personality, team perfor-mance, and organizational context. In P. Whitney & R. B. Ochsman(Eds.), Psychology and prod~tivity (pp. 93-103). New York: Plenum.

Holland, J. L. (1985). Making vOcational choices: A theory of careers.Englewood Cliffs, NJ: Prentice-Hall.

Hollander, E. P., & Julian, J. W. (1969). Contemporary trends in theanalJrsis of leadership processes. Psychological Bulletin, 7 J. 387-9 I.

Hough, L. M. (1992). The "Big Five" personality variables-constructconf~lSion: Description versus prediction. Human Peiformance. 5. 139-155.

House, R. J. (1977). A 1976 theory of charismatic leadership. In J. G.Hun1 & L. L. Larson (Eds.), Leadership: The cutting edge (pp. 189-207). Carbondale: Southern Illinois University Press.

House, R. J., Spangler, W. D., & Woycke, J. (1991). Personality andcharisma in the U. S. presidency: A psychological theory ofleadershipeffectiveness. Administrative Science Quarterl}: 36. 364-396.

Howard, A., & Bray, D. W. (1990). Predictions of managerial successover J,>ng periods of time: Lessons for the Management Progress Study.In K. E. Clark & M. B. Clark (Eds.), Measures of leadership (pp.113-][30). West Orange, NJ: Leadership Library of America.

Howell, J. M., & Frost, P. (1988). A laboratory study of charismatic

leadelship. Organizational Behavior and Human Decision Processes.43. 2'~3-269.

Hughes, R. L., Ginnett, R. A., & Curphy, G. J. (1993). Leadership:Enhancing the lessons of experience. Homewood, IL: Irwin.

Johnston, W. B, & Packer, A. H. (1987). Workforce 2000. Indianapolis,IN: Hudson Institute.

Kanter, Il M. (1983). The change masters. New York: Simon & Schuster.Kaplan, R. E., Drath, W. H., & Kofodimos, J. R. (1991). Beyond am-

bition' How driven managers can lead better and live bettel: San Fran-

cisco: Jossey-Bass.

Kenny, D. A., & Zaccaro, S. J. (1983). An estimate of variance due totraits inieadership. Journal of Applied Psychology. 68. 678-685.

Lombardo, M. M., Ruderman, M. N., & McCauley, C. D. (1988). Ex-planations of success and derailment in upper-level management p0-sitions. Journal of Business and Psychol~ 2, 199-216.

Lord, R. G., De Vader, C. L., & Allinger, G. M. (1986). A meta-analysisof the relationship between personality traits and leadership percep-tions: An application of validity generalization procedures. Journalof Applied Psychology. 7 J, 402-410.

Lord, R. G., Foti, R. J., & De Vader, C. L. (1984). A test of leadershipcategoJization theory: Internal structure, information processing, andleaden:hip perceptions. Organizational Behavior and Human Perfor-mance, 34, 343-378.

Mann, R. D. (1959). A review of the relationships between personalityand performance in small groups. Psychological Bulletin. 56. 241-270.

McCall, ,\11. W., & Lombardo, M. M. (1983). Off the track: Why andhow successful executives get derailed (Tech. Rep. No. 21). Greensboro,NC: Cl:nter for Creative Leadership.

McClellaod, D. C. (1975). Power: The inner experience. New York: Ir-

vington.McClella.l1d, D. C., & Bumham, D. (1976). Power is the great motivator.

Harvard Business Review. 25, 159-166.McCrae, R. R., & COSta, P. T., Jr. (1987). Validation of the five-factor

model of personality across instruments and observers. Journal ofPerson.7/ity and Social Psychology. 52. 81-90.

McEvoy, G. M., & Beatty, R. W. (1989). Assessment centers and sub-ordinal:e appraisals of managers: A seven-year examination of pre-dictive validity. Personnel Psychology. 42. 37-52.

Meindl, J. R., & Ehrlich, S. B. (1987). The romance of leadership andthe evaluation of organizational perfonnance. Academy of Manage-ment Journal, 30. 91-109.

Millikin-[>avies. M. (1992). An exploration offlawedfirst-line supervisionUnpublished doctoral dissertation. University of Tulsa, Tulsa, OK.

Mintzber)' H. (1982). If you're not serving Bill or Barbara, then you'renot 5er,ring leadership. In J. G. Hunt, U. Sekaran, & C. A. Schriesheim(Eds.), Leadership beyond establishment views (pp. 239-259). Car-bondall:: Southern Illinois University Press.

Morrison, A. M. (1992). The new leaders: Guidelines on leadership di-versity in America. San Francisco: Jossey-Bass.

502 June) 994 .American Psychologist

Morrison, J. D. (1993). Group composition creative performance,Unpublished doctoral dissenation, Univers ty of Tulsa, Tulsa, OK,

Murphy. K. R.,&Oeveland,J, N. (1991). Perfo nceappraisal, Boston:Allyn & Bacon.

Nilsen, D. (1992). Using observers' judgm for selection. In R. T.Hogan (Chair), The future of leadership se ection. Symposium con-ducted at the 13th Biennial Psychology. the DoD Conference,U. S. Air Force Academy, Colorado Sprin CO.

Nilsen D., & Campbell, D. P. (1993). Self-o er rating discrepancies:Once an overrater, always an overrater? Hum Resource Management.32.265-281.

Norman, W. T. (1963). Toward an adequate onomyofpersonalityattributes: Replicated factor structures in nomination personalityratings. Journal of Abnormal and Social P hology, 66. 574-583.

Offermann, L. R., & Gowing, M. K. (1990). izations of the future.American Psychologist. 45. 95-108.

Passini, F. T., & Norman, W.l: (1966). A un' conception of per-sonality structure? Journal of Personality Social Psycholog}: J.44-4 9 .

Peabody, D., &. Goldberg, L. R. (1989). Sam determinants of factorstructures from personality-trait descripto .Joumal of Personalityand Social Psychology, 57. 552-567.

Personnel Decisions, Inc. (1991). The PROFl ~. Minneapolis, MN:Author.

Peters, T. J., & Waterman, R. H. (1982). In s arch of excellence. NewYork: Harper & Row.

Peterson, D. B. (1993). Measuring change: A ychometric approach toevaluating individual training outcomes. I V. Arnold (Chair), In-novations in training evaluation: New me ~s. new designs, Sym-posium conducted at the Eighth Annual C nference of the Societyfor Industrial and Organizational Psychol , San Francisco.

Peterson, D. B., & Hicks, M. D. (1993, ). How to get people tochange. Workshop presented at the Eighth nnual Conference of theSociety for Industrial and Organizational chology, San Francisco.

Pulakos, E. D., Borman, W. C., & Hough, L. .(1988). Test validationfor scientific understanding: Two dem .ons of an approach forstudying predictor-criterion linkages. Perso el Psycholog}: 41. 703-716.

Ross, S. M., & Offermann, L. R. (1991, April). ransformational leaders:Measurement of personality attributes and rk group performance,Paper presented at the Sixth Annual Conti ce of the Society forIndUStrial and Organizational Psychology, S Louis, MO.

Ropp, K. (1987, February). Restructuring: S .of the fittest. Per-sonnel Administrator, pp. 45-47.

Rueb, J. D., & Foti, R. J. (1990, April). TI; is. selfmonitoring. andleadership emergence. Paper presented at the ifth Annual Conferenceof the Society for Industrial and Organizati Psychology, Miami,FL.

Rush, M. C., Thomas, J. C., & Lord, R. G. (I 77). Implicit leadershiptheory: A potential threat to the internal idity of leader behaviorquestionnaires. Organizational Behavior Human Performance.20.93-110.

Schneider, B. (1987). The people make the plal::e. Personnel Psycholo~40. 437- 453.

Shannon, C., & Weaver, W. (1949). The mathematical theory of com-munication. Urbana: University of Illinois Press.

Shippex; F., & Wilson, C. L. (1991, July). The impact of managerialbehaviors on group performance. stress. and commitment. Paper pre-sented at the ImpaCt of Leadership Conference, Center for CreativeLeadership, Colorado Springs, CO.

Smith, J. E., ~n, K. P., & Alexander, R. A. (1984). Leadership: Itcan make a difference. Academy of Manag~ment Journal. 27. 765-776.

Sorcher, M. (1985). Predicting executive success: What it takes to makeit in senior management. New York: Wiley.

Snyder, M. (1974). The self-monitoring of expressive behavior. JournalQf Personality and Social Psychology. 30. 526-537.

Stogdill, R. M. (1948). Personal faCtors associated with leadership: Asurvey of the literature. Journal of Personality. 25,35-71.

Stogdill, R. M. (1974). Handbook of leadership. New York: Free Press.Sweetland, J. (1978). Work in America Institute studies in productivity:

Highlights of the literature: Managerial productivity. Scarsdale, NY:Work in America Institute.

Tellegen. A. (1982). Brief manual for the Multidimensional Personalit}.Questionnaire (mimeograph). (Available from A. Tellegen, Depanmentof Psychology, University of Minnesota, Minneapolis, MN 55455)

Tellegen, A. (1985). StruCtures of mood and personality and their rele-vance to assessing anxiety, with an emphasis on self-report. In A. H.Truman & J. D. Maser (Eds.), Anxiety and anxiety disorders (pp.681-706). New York: Erlbaum.

Ten, R. P., Jackson, D. N., & Rothstein, M. (1991). Personality measuresas prediCtors of job performance: A meta-analytic review. PersonnelPsycholo~ 44. 703-742.

Tupes, E. C., & Christal, R. E. (1961). Recu"ent personality factorsbased on trait ratings (ASD-TR-61-96). Lackland Air Force Base,TX: Aeronautical Systems Division, Personnel Laboratory.

Van Veisor, E., Taylor, S., & Leslie, J. B. (1992, AI1gust). Self/rater agree-ment, sl~/f-awareness and leadership effectiveness. Paper presented atthe lOOth Annual Convention of the American Psychological Ass0-ciation, Washington, DC.

Weiss, H. ]~., & Adler, S. ( 1981). Cognitive complexity and the struCtureof implicit leadership theories. Journal of Applied Psychology, 66, 69-78.

Yuki, G. A. (1989). Leadership in organizations (2nd ed.). EnglewoodCliffs, NJ: Prentice-HaIl.

Yuki, G. A., Wall, S., & Lepsinger. R. (1990). Preliminary report on thevalidation of the management practices survey. In K. E. Clark &M. B. Oark (Eds.), Measures of leadership (pp. 223-238). West Orange,NJ: Le2dership Library of America.

Zaccaro, ~.. J., Foti, R. J., & Kenny, D. A. (1991). Self-monitoring andtrait-ba.~ variance in leadership: An investigation of leader flexibilityacross fl1uJtiple group situations. Journal of Applied Psychology. 76,308-315:

I APPENDIX

The ~ig Five Dimensions oi: Personality

SurgencySurgency measures the degree to hich an individualis sociable, gregarious, assertive, a d leaderlike versusquiet, reserved, mannerly, and with rawn. Some of themore common personality traits sociated with thisdimension include dominance, ca acity for status, orsocial presence (Goug;b, 1987), e need for power(McClelland, 1975), sociability (R Hogan & Hogan,

1992), surgency (Tupes & Christal, 1961), or assertive-ness (Borgatta, 1964).

Emotional StabilityThis dimension of personality concerns the extent to whichindividuals are calm, steady, cool, andself-conndent versusanxious, insecure, worried, and emotional. Some of the per-sonality traits associated with emotional stability includeneurotic:ism (Eysenck, 1970; McCrae & Costa, 1987), emo-

503June 1994 .American Psychologi~

versus grumpy, unpleasant, disagreeable, and cold. Per-sonality traits associated with this dimension includelikeability (Borgatta, 1964; R. Hogan & Hogan, 1992),friendly compliance (Digman, 1988), need for affiliation(McClelland, 1975), and love (Peabody & Goldberg,1989).

Inteillectance

This i:limension of personality concerns the extent towhich an individual is imaginative, cultured, broadminded, and curious versus concrete minded, practical,and hclS narrow interests. Personality traits associated withthis dimension include culture (Norman, 1963; Tupes &Christal, 1961) and openness to experience (McCrae &Costa, 1987).

tiona! stability (Guilford, 1975), negative affectivity (Telle-geri, 1985), and affect (Peabody & Goldberg, 1989).

ConscientiousnessConscientiousness differentiates individuals who are hard-working, persevering, organized, and responsible fromthose who are impulsive, irresponsive, undependable, andlazy. Personality traits categorized under this dimensioninclude prudence and ambition (R. Hogan & Hogan,1992), will to achieve (Digman, 1988), need for achieve-ment (McClelland & Burnham, 1976), dependability(Tripes & Christal, 1961), constraint (Tellegen, 1985), andwork (Peabody & Goldberg, 1989).

AgreeablenessAgreeableness measures the degree to which individualsare sympathetic, cooperative, good-natured, and warm

I