Mehra Kilduf Brass 2001 Self Monitors

8/12/2019 Mehra Kilduf Brass 2001 Self Monitors

1/27

The Social Networks ofHigh an(d Low Self-monitors; Implicationsfor WorkplacePerformance

Ajay MehraUniversity of CincinnatiMartin KilduffThe Pennsylvania StateUniversityDaniel J. BrassUniversity of Kentucky

2001 by Cornell University.0001-8392/01/4601-0121/$3.00.

We thank Dennis Gioia, Giuseppe (Joe)Labianca. Marianne Lewis, James Martin,Suzanne Masterson, Hongseok Oh, KevinSteensma, Charles Trevof, Wen pin Tsai,and three anonymous SQ reviewers forsuggestions and comments on previousdrsits. We are also grateful to Reed Nel-son for editorial guidance, to Linda Johan-son for helpful copy editing and lo RonButt, David Krackhardt, and Mark Snyderfor en courageme nt. Earlier versions of

This article examines how different personality types crate and benefit from social networks in organizations.Using data from a 116-member high-technology f i rm , w etested how self-monitoring orientat ion and network posi-t ion related to work performance. First , chameleon-likehigh self-monitors were more l ikely than true-to-them-selves low self-monitors to occupy central positions insocial netw orks. Secon d, for h igh (but no t for low ) selfmonitors, longer service in the organization related to toccupancy of strategically advantageous network posi-t ions. Third, self-monitoring and centrality in social neworks independently predicted individuals ' workplaceperformance. The results paint a picture of people shap-ing the networks that constrain and enablepe r fo rmance .*

One of the enduring questions we face as human beingsconcems why some people outcompete others in the racefor life's prizes. In work organizations, for example, why asome people better performers than others? One answer tothis question is provided by research on the innportance ofstructural position. Within each specific work context, somindividuals occupy more advantageous positions in social nworks than other individuals. These positions allow access people who are otherwise disconnected from each other.The individuals who act as go-betw een s, bridging the strutural ho les betw een disconnected others, facilitate resourflow s and know ledge sharing across the organization. Theicontributions to organizational functioning may lead to

enhanced rewards, including faster promotions (Burt, 1992and higher performance ratings.

Research on structural position has emphasized the impor-tance of being in the right place (Brass, 1984) but has negleed both the possibility that the network positions occupied individuals might be influenced by their psychology and thpossibility that personality and social network position migcombine to influence important outcomes such as work peformance. The structural approach to organizational dynamitends to emphasize the structure of positions in social space(Pfeffer, 1991; Blau, 1993) and avoids dependence on difficuto-measure psychological properties of actors (e.g., McPher-son, Popielarz, and Drobnic, 1992). Recent calls for moreinsight into the origins of network positions and the impor-tance of individual characteristics (e.g., Emirbayer and Good-win, 1994) prompt us to investigate why some individualsoccupy structurally advantageous positions and how individdifferences in psychology and structural position combine determine performance in organizational contexts.

The structuralist approach is not alone in disregarding thepossible effects of individual characteristics on social structures. Despite a long history of psychological research suggesting that individuals differ with respect to social influe


2/27

that both enable and constrain action. We follow in the tradi-tion of those who recognize the importance of understandingthe micro-foundations of structural patterns (e.g., Granovet-ter, 1973; Ibarra, 1993; Uzzi, 1996).

Earlier work by social network pioneers included personalitymeasures (e.g., New com b, 1961 ; Sampson, 1968) and inter-personal orientations (e.g., Breiger and Ennis, 1979; see alsorecent work by Janicik, 2000). In bringing the individual backinto social network analysis (cf. Kilduff and Krackhardt, 1994),we build on this previous work. Rather than treat individualattributes and social attributes as separate realms of enquiry,we seek to understand how the social networks that signifi-cantly affect the performance of organizational participantsare shaped by the attributes of interacting individuals.

THEORY

The Structural AdvantageIndividuals may outperform their peers because of differ-ences in the networks to which they belong. Links to friendsand work partners can provide the assistance and social sup-port necessary for high performance, but not all network con-figurations are likely to be equally helpful. Forming a largenetwork, for example, may be less important than acquiring astructurally advantageous position within a network (Burt,1992).

Social actors who connect disconnected others tend to gainboth information and control benefits. Information concerningprojects, crises, resources, and other contingencies flowfrom a diversity of social actors to the central actor whoseties link disconnected others. Actors whose social ties arelimited to one clique are less likely to receive diverse infor-mation than are actors whose ties span cliques becauseinformation that circulates within a clique of highly connectedworkers is likely to be redundant. Evidence for the benefitsof structural holes comes from both small-group and organi-zational research (see the reviev^ in Burt, Jannotta, andMahoney, 1998). Small-group experiments showed that peo-

ple with exclusive relations to otherwise disconnected con-tacts tended to gain greater resources (Cook and Emerson,1978; Cook et al., 1983). One organizational study examinedthe importance for non-supervisory personnel of occupyinghigh-betweenness centrality positionsthat is, positions thatenable occupants to act as potential go-betweens for thosenot connected with each other. Results showed that thehigher the betweenness centrality in the informal comm uni-cation network, the greater the social influence and the high-er the likelihood of promotion to supervisor within the follow-ing three-year period (Brass, 1984).

Occupying a position betw een disconnected others is impor-tant not only for non-supervisory personnel but also for those


3/27

Networks and Self-monitoring

works that tied them to unconnected others tended to havehigh mobility (Podolny and Baron, 1997).

The accumulating evidence suggests that individuals withties across social divides gain non-redundant information concerning opportunities and resources. The ability to obtainresources such as information is directly related to individualand group performance (O'Reilly, 1977; O'Reilly and Roberts1977a, 1977b). Further, actors who connect disconnectedothers can facilitate the flow of infornnation across the wholsystem of coordinated activity that nnakes up the organiza-tion, thereby contributing to the accomplishment of organization-wide goals. Given this, when we discuss individual per-formance in this article, we refer to the extent to whichindividuals contribute to organizational purposes, building onthe work in organization theory that emphasizes that job per-formance consists of individuals contributing to the tasksspecific to the organization (Burns and Stalker, 1994: 97). Pre

vious work has focused on the effects of structural positionon outcome variables such as power and promotions but hasoffered little conclusive evidence concerning performance inorganizations. One of the few studies that did examine workperformance found that employees occupying central posi-tions in the workflow network were no more likely to be hiperformers than employees occupying less central positions(Brass, 1981). In contrast, research on officers and enlistedmen in three high-technology military organizations showedthat people with two or more network contacts performedbetter than people with one or no network contacts (Roberts

and O'Reilly, 1979). This research did not examine the importance of network centrality or ties that link disconnected others. Given these suggestive but inconclusive findings, it isuseful to examine directly whether structural position pre-dicts workplace performance.

Self-Monitor ing

Individuals in organizations may outperform their peers notonly because of differences in the networks to which theybelong but also because of individual differences in personalty. Of the many personality variables that could potentiallyaffect performance, self-monitoring, a variable centrally concerned w ith individuals' active construction of public selveto achieve social en ds (Gangestad and Snyder, 2000: 546),stands out for three reasons. First, self-monitoring theoryprovides compelling arguments linking individual differencein self-monitoring with a range of job outcomes, such as performance in the workplace, leadership emergence in workgroups, conflict management, information management,impression management, and boundary spanning (Snyder,

1987: 88-90; Kilduff and Day, 1994). Second, self-monitorintheory makes clear predictions concerning the effects of selfmonitoring orientation on how individuals shape social worl


4/27

According to self-monitoring theory, individuals differ in theextent to which they are willing and able to monitor and con-trol their self-expressions in social situations. Some peopleresemble successful actors or politicians in their ability to findthe appropriate words and behaviors for a range of quite dif-ferent social situations. With chameleon-like ease, they pre-sent the right image for the right audience. Other people, by

contrast, appear to take to heart the advice Polonius gave toLaertes in Shakespeare's Hamlet, To thine ow n self bet rue : they insist on being themselves, no matter how incon-gruent their self-expression may be with the requirements ofthe social situation. Research on self-monitoring has providedimportant insights into individual differences in how individu-als present themselves in social contexts (see Gangestadand Snyder, 2000, for a review ).

In a social situation, high self-monitors ask, W ho does thissituation want me to be and how can I be that person?(Snyder, 1979). By contrast, low self-monitors ask, W ho amI and how can I be me in this situation? (Snyder, 1979; Kil-duff and Day, 1994). Self-monitoring theory, therefore, pro-vides new insight into the age-old question of whetherbehavior is a function of consistent dispositions or strong sit-uational pressures. From a self-monitoring perspective, someindividuals (the low self-monitors) are consistent in demon-strating behavior that expresses inner feelings, attitudes, andbeliefs. Other individuals (the high self-monitors) are consis-tent in adjusting behavior to the demands of different situa-tions.

Because high self-monitors rely on social cues from others toguide their behaviors rather than on their own inner attitudesand emotions, high self-monitors are more likely than lowself-monitors to resolve conflicts through collaboration andcompromise (Baron, 1989)- Further, high self-monitors tendto emerge as group leaders (Zaccaro, Foti, and Kenny, 1991),particularly in situations calling for high levels of verbal inter-action (Garland and Beard, 1979) and in normative climatesthat support the emergence of leadership (Whitmore andKlimoski, 1984).

High self-monitors tend to emerge as leaders perhaps In partbecause they are more skilled at social interactions (Furnhamand Capon, 1983). One study found that low self-monitorsattended more to internal cues to produce effective work,whereas high self-monitors attended to situational cues,including the leadership behavior of supervisors (Andersonand Tolson, 1989). High self-monitors are more active in con-versations (Ickes and Barnes, 1977) and tend to talk aboutthe other person (and other people) instead of talking aboutthemselves (Ickes, Reidhead, and Patterson, 1985). High self-monitors are better than low self-monitors at pacing conver-sations (Dabbs et al.. 1980), using humor (Turner, 1980), and


5/27

Networks and SeK monrtoring

The social skills and leadership abilities of high self-monitors,therefore, may enable them to perform significantly betterthan low self-monitors in the modern workplace where coop-eration with others to achieve organizational purposes is thenorm and where leadership emergence is encouraged (seereview by Baron and Markman, 2000). Although there is noreason to suppose that self-monitoring orientation affects the

proficiency with which individuals perform technical duties,contextual activities, such as cooperating with others and fol-lowing procedures even when they are personally inconve-nient, are also a major part of workplace performance (Bor-man and Motowidio, 1993). Much managerial work involvescom mu nicating w ith others (Gronn, 1983), performing a vari-ety of different roles (Mintzberg, 1973), and relating to theneeds of a large number of diverse people (Kotter, 1982). Thesocial skills and leadership abilities characteristic of high self-monitors may enable them to perform better than low self-monitors in such contexts.

Previous research has shown that individual differences inhow people approach social situations affect individual attainment in managerial careers. Self-monitoring effects havebeen demonstrated on managerial promotions over a five-year period: high self-monitors are more likely to be promot-ed in managerial careers than low self-monitors (Kilduff andDay, 1994). Much of the pioneering work concerning theeffects of self-monitoring on performance-related variableshas consisted of laboratory studies on students (e.g., Cald-wel and O'Reilly, 1982a). The occasional field study has tended to focus either on the eventual outcome s of performancedifferences, such as early promotions (e.g., Kilduff and Day,1994), or has focused on specific types of workers, such asboundary spanners (e.g., Caldwell and O'Reilly, 1982b). It isimportant, therefore, to test wh ether self-monitoring predictworkplace performance across the full range of organizationapositions in an organization.

Three Models

Given the separate and unrelated literatures on social net-works and personality, the question is how structural positionand self-monitoring combine to affect individual performancin organizations. We explore three perspectives: a mediationmodel, an interaction model, and an additive model.

Mediation model. Performance differences among individu-als in organizations may be due to the tendency of a particu-lar personality type (the high self-monitor) to occupy struc-turally central positions that link otherwise disconnectedpeople and provide differential resources. Research across arange of social relationships shows that high and low self-monitors tend to inhabit different social worlds (Snyder,Gangestad, and Simpson, 1983; Snyder and Simpson, 1984;S d Si d G g t d 1986) Abl t t il b h


6/27

It is also possible to formulate a model inwhich self-monitoring mediates theeffects of struclural position on perfor-mance. Such a model would imply thatthe individual's seif-monitoring orientationoould be changed by the occupation of a

central position. This idea has no supportm self-monitoring theory or self-monitor-ing evidence (as summarized in the me th-ods section) and found no support in sup-

The high self-monitor likes to have one friend for tennis,another friend for basketball, and yet another friend forchess. High self-monitors maintain flexibility and make littleemotional investment in relationships. Friends are chosenbased on how closely their skills match activity domains. Asone high self-monitoring tennis player observed , W hen Iwant to play tennis, I select a partner who can challenge me

(quoted in Snyder, 1987: 65). Low self-monitors, by contrast,tend to choose friends on the basis of liking, irrespective ofwhether the friends are proficient in tennis, basketball, orchess. They like to be with the same friends across activitydomains (Snyder, Gangestad, and Simpson, 1983). As onelow self-monitor commented about her choice of an activitypartner, Jan's my best friend. Besides, she's the most funto be around, whatever the ac tivity (quoted in Snyder 1987-65).

Self-monitoring theory predicts, therefore, that high self-mon-

itors, relative to low self-monitors, will tend to develop friendship relations at work with distinctly different people. Where-as low self-monitors will tend to occupy relativelyhomogenous social worlds, high self-monitors w ill tend todevelop relationships across groups, using their flexible iden-tities to play different roles in different groups. In a work-place, high self-monitors are therefore likely to bridge socialworlds, acting as conduits through which otherwise uncon-nected people exchange information.

According to the mediation perspective, high self-monitors

will occupy central positions in social networks in organiza-tions and reap the benefits of access to diverse resourceflows and information detailed by structural sociologists (e.g.,Burt, 1992). Because they tend to serve as go-betweensbetween disconnected others, high self-monitors willenhance their value to the organization by facilitatingresource flows and knowledge sharing across the organiza-tion and thereby achieve superior performance. Thus, highself-monitors will tend to perform better than low self-moni-tors as a direct result of their differential success in occupy-ing structurally advantageous positions in social networks.

Complete mediation would suggest that any effect of self-monitoring on work performance is due to the individual'sstructural position in social networks. Complete mediation,therefore, would offer some support for the structuralist view(e.g., B urt, Jannotta, and Mahoney, 1998) that individual dis-positions can serve as proxies for the network positions thatindividuals are likely to occupy.'

Interaction model. The different, but not incompatible, inter-action perspective suggests that different personality typesmay differentially take advantage of structural positions. High

self-monitors may be more able and motivated than low self-monitors to seek out and use the resources available fromthe different social groups accessible from bridging positions


7/27

Networks and S elf-monitoring

advantage of the opportunities represented by such posi-tions. The interaction model suggests that both a high self-monitoring disposition and a structurally advantageous posi-tion in the social network are necessary for the individual toachieve high work performance.

Numerous studies have confirmed that high self-monitors,compared with low self-monitors, tend to be more respon-sive to the specific characteristics of situations (see review inSnyder, 1987: 33-46}. For example, in one study, high self-monitors showed themselves acutely sensitive to the differ-ing contexts in which social interaction took place. The highself-monitors were conformist in social situations in whichconformity was the most appropriate interpersonal orienta-tion and were nonconformist when reference group normsfavored autonomy. By contrast, low self-monitoring groupmembers were virtually unaffected by their social settings(Snyder and Monson, 1975). This differential responsivenessis likely to affect work performance. In a field study of peopwhose iobs required that they interact with groups whosenorms differed from one another, high self-monitors outper-form ed low self-monitors (Caldwell and O Reilly, 1982b}. Thstudy, wh ich focused on wo rke rs links outside the organiza-tion, provides support for the interaction model. Extendingthis research to the current study of workers within the organization, we might expect to find that only high self-monitorare able to take advantage of structurally advantageous net-work positions to enhance performance.

A further reason to expect performance differences for highand low self-monitors occupying bridging positions relates tothe detection of useful social information. High self-monitorare better at scanning the social world for information aboutpeople and their intentions. High self-monitors are more likethan low self-monitors to notice and remember informationconcerning others (Berscheid et al., 1976}, to be more suc-cessful at detecting people s intentions (Jones and Baumeis-ter, 1976}, and to be more accurate at eyewitness identifica-tion (e.g., Hosch et al., 1984}- If valuable information isavailable to those occupying bridging positions in social net-works, then it is more likely to be detected by high self-montors than by low self-monitors.

Additive model. We have argued that high and low self-monitors may differentially succeed in organizations becausethey differentially occupy structurally advantageous positionin social networks (the mediation perspective) or becausehigh self-monitors may be differentially able to capitalize onstructurally advantageous positions (the interaction perspec-tive), A third possibility is that structural position and self-monitoring may have relatively independent, additive effectson performance in organizations. The additive model involvetwin predictions concerning work performance. The structurposition prediction is that the greater the extent to whichi di id l i l b f h


8/27

Figure 1. Three models of how self monitoring and structural position affect individual performance inorganizations.

MEDIATION MODEL

|Self-Monitoring | Structural Position Performance

INTERACTION MODEL

Self-Monitoringl

Structural Position Performance

ADDITIVE MODEL

tages in work performance: (a) occupy a structurally advanta-geous network position; (b) possess a high self-monitoringorientation.

Figure 1 sumnnarizes the three modeis of the possibie

effects of structural position and seif-monitoring on perfor-mance that we tested in our study.

METHOD

SiteBayou Corporation (a pseudonym) was a small high-technoio*gy company invoived in the chemical analysis of complexcomp ounds. E mployees researched, produced, and m arketedhigh-precision chromatographic equipment for laboratoriesand other clients that analyzed the chemical composition offoods, fragrances, petrochemicals, Pharmaceuticals, andother products. Bayou was founded in 1985 by an entrepre-neur who left his job at a medium-sized chemical company to


9/27

Networks and Self monitoring

nization had created an entrepreneurial culture that ennpha-sized informality rather than bureaucracy.

Bayou competed in fast-moving markets against much largercompanies such as Hewlett-Packard. The company founderemphasized the importance of innovation and creativity asthe keys to survival in this competitive marketplace. Organi-zational structure v\/as kept deliberately flat, with only threelevels of hierarchy. Instead of departments, employees wereorganized into fluid workgroups that ranged in size from twoto sixteen. The company prided itself on being in the fore-front of equal opportunity employment and had won awardsfor its success in recruiting and promoting women.

Data

We collected network and personality data by means of aquestionnaire sent to all 116 employees (68 men and 48wo m en). W e collected performance-rating data by means of

a separate questionnaire sent to all 22 supervisors (17 menand 5 women). Data about reporting relationships, demogra-phy, and tenure came from company records.

The response rate was 88 percent for the questionnaire sentto all employees and 100 percent for the questionnaire sentonly to supervisors. Non-respondents did not differ signifi-cantly from respondents with regard to sex, tenure, or performance. Missing data on self-monitoring reduced the usablesample from 102 to 93 individuals for analyses involving thisvariable. Because there were no performance measures for

the head of the company, analyses concerning both perfor-mance and self-monitoring used a sample of 92,

Measures

Social networks. We collected data on friendship relationsand workflow relations using the roster method. For eachnetwork, we asked respondents to look down an alphabeticallist of employees and place checks next to the names of peo-ple they considered friends or work partners. Data for eachrelation were arranged in 102 x 102 binary matrices, In eachmatrix, cell X^ corresponded to i's relation to j as reported byi. For exam ple, if i reported j as a friend , then cell X^ in thefriendship ma trix w as coded as 1, othe rwise X^ w as coded a0. Each matrix contained 10,302 observations on all possiblepairs of people.

For each network question, respondents were free to nomi-nate as many network contacts as they deemed appropriate.This format is preferable to a fixed-choice design in whichrespondents are asked, for example, Lis t your four bestfrie nd s, because it is unlikely that all people have exactlyfour best friends. Limiting respondents to a fixed number ofchoices tends to introduce measurement error into networkdata (Holland and Leinhardt, 1973).


10/27

nected. Research has shown that individuals are reliablesources of information concerning the membership of stablenetworks to which they themselves belong (Freeman, Rom-ney, and Freeman, 1987), but ego's responses concerningpossible interconnections between people to whom ego istied are subject to systematic bias (Kumbasar, Romney, andBatchelder, 1994; Krackhardt and Kilduff. 1999). Thus, ego-network data used to assess structural holes are potentiallydisto rted by perceptual biases. i

Comparing workflow and fr iendship networks. Asresearch on social networks has pointed out (e.g., Roethlis-bergerand Dickson, 1939: 493-510), in considering theimportance of network position in an organization, two typesof networks need to be considered: the workflow and infor-mal networks. The workflow network is the formally pre-scribed set of interdependencies between employees estab-lished by the division of labor in the organization. Work flowsthrough the organization as workers exchange inputs andoutputs. A successful interaction in the workflow networkenables the flow of work from one person to another (Brassand Burkhardt, 1992: 197).

By contrast, informal social networks, such as the friendshipnetwork, derive from mutual liking, similarity of attitudes, orpersonal choice. Compared w ith the w orkflo w network, thefriendship network represents more individual choice and ini-tiative. People have more discretion in the choice of friendsthan they have in the choice of with whom to interact toaccomplish work. Achieving a structurally advantageous posi-tion in either the more formal workflow network or the moreinformal friendship network can bring benefits to the individ-ual in terms of diverse information and other resources.

Friendship network Respondents were asked to look downan alphabetical list of fellow employees and place checksnext to the names of those individuals they considered especially good frien ds. Friends we re defined as peop lewith whom you like to spend your free time, people you havebeen with most often for informal social activities, such as

visiting each other's homes, attending concerts or other pub-lic performances.

Workflow network was modeled after Brass (1981: 332),w ho argued that task positions and the workers occ upyingthese positions [can be] viewed as interrelated on the basisof the flow of work through the organization. Respondentswere asked to place a check next to the names of their work-flow contacts. We combined workflow inputs and workflowoutputs to make the questionnaire more manageable andbecause Brass (1984) found no differences between the pre-

dictive power of input and output contacts. Workflow con-tacts were d efined as the se t of people that provide youwith your workflow inputs taken together with the set of


11/27


organization rather than in the more discretionary task advicenetworks studied by others (e.g., Podolny and Baron, 1997).

Network size and structure A large network, one withmany contacts, can enable the individual to access numerousothers for information and other resources. But the benefitsof a large network may be offset by the costs involved in

maintaining a large number of relationships (Rook, 1984).People who interact with numerous others in organizationsrun the risk of running short of time and other resources necessary for work performance. Thus, people with large net-works within the organization may not necessarily achievethe highest performance ratings. They may be so busy main-taining ties at work that their work performance suffers (seeBurt and Ronchi, 1990, fora case study). In considering hownetwork position relates to work performance, it is thereforeimportant to examine simultaneously the relationshipsbetween network size and performance and between

betwee nness centrality and performance. One of the ques-tions that our research attempts to answer is, controlling forthe size of the individual s netw ork, does the exten t to w hicthe individual s netwo rk spans social divides predict wo rk-place performance? By looking at both network measuressimultaneously, we can assess whether network size andnetwork betweenness have independent relationships withwork performance.

Betweenness centrality As a measure of the extent to whicheach individual occupied a structurally advantageous position

connecting otherwise unconnected others in the friendshipand workflow networks, we used betweenness centrality(Freeman, 1979). We chose this measure rather than a morelocal measure of autonomy, such as constraint (Burt, 1992),because betweenness centrality takes both direct and indi-rect ties into account (Brass, 1984; Krackhardt, 1990; Brassand Burkhardt, 1993), whereas constraint focuses primarilyon the direct ties in ego s im mediate circle of contacts. Mo rlocal measures of the extent to which individuals span struc-tural holes are useful when sampling from large populationsfor which whole network data are unavailable (e.g, Burt,1992).The (102 X 102) friendship matrix and the (102 x 102) work-flow matrix were each submitted to the betweenness proce-dure in the network program UCINET IV (Borgatti, Everett,and Freeman, 1992: 85; see Freeman, 1979, for the formula)The higher the betweenness score of an actor, the greaterthe extent to which that actor serves as a structural conduitconnecting others in the network. More formally, between-ness centrality measures the frequency with which an actorfalls between other pairs of actors on the shortest or geodes-

ic paths connecting them (Freeman, 1979: 221).Because it is difficult to interpret measures of betweenness


12/27

member of the pair nominated the other. The pattern ofresults remained unchanged.

Network size was mea sured as the total num ber of each indi-vidual's direct links with other actors in the network, a mea-sure also kno wn as degree cen trality (Scott, 19 91: 86- 87). Tobe compatible with measures of betweenness centrality, we

calculated size on friendship and workflow matrices sym-metrized according to the rule that if either member of a pairnom inated the other, the pair wa s con sidered to have a tie.

Performance Our theory of job performance emphasizes theextent to which individuals succeed (in the eyes of manage-ment) in contributing to organizational ends. In the absenceof objective measures of performance across job types inthis organization, we relied on supervisory ratings. Using a 6-item scale arranged in 5-point Likert format, supervisors ratedthe performance of those subordinates who reported directlyto the m . As researchers have noted , in work organizations the vast majority of performance ratings come directly fromthe imme diate supervisor (Bretz, Milkovich, and Read, 1992:3 3 1 ; see also Scullen and Mount, 2000). A recent compre-hensive review of performance evaluation in work settingsconclud ed that supervisory ratings are mos t likely validreflections of true performa nce (Arvey and Murphy, 1998:163).

We informed supervisors that performance ratings would beconfidential and used only for research purposes. Perfor-mance ratings obtained for research purposes tend to bemore reliable and valid than those obtained for administrativepurposes (Wherry and Bartlett, 1982). The six performanceitems were selected after extensive discussions with thefirm's human resource director and a group of four employ-ees represen ting a range of job types at the f irm.

Supervisors first evaluated subordinates' performance onthese three items : (1) the overall job performance of theindividual (1 = poor, 5 = exce llent); (2) the likelihood that thesubordinate wo uld achieve future career related success(such as prom otions, awa rds, bonuses, and involveme nt inhigh-profile pro jects) at Bayou (1 = very unlikely, 5 = verylikely); and (3) the likelihood tha t you wo uld pick [the subor-dinate] to succeed you in your jo b (1 = very unlikely, 5 =very likely). \

Given the strong emphasis placed on innovation at Bayouand the growing recognition among researchers of the impor-tance of contextual aspects of job performance (e.g., Bormanand Motowildo, 1993; Arvey and Murphy, 1998), we also

included three items, taken fro m Scott and Bruce (1994), tocapture employees' workplace innovativeness. Supervisorsrated subordinates' innovativeness (using 5-point scales) on


13/27


The reliability of the six-item scale, as measured by Cron-bach's (1951) alpha, was .90. The results of a componentanalysis showed all six items loaded on the same component(eigenvalue = 4.06; all loadings were above .76) thatexplained 68 percent of the overall variance. To checkwhether our results were an artifact of the composition ofour performance measure, for all analyses that included per-formance, we ran separate tests using (1) the final six-itemmeasure of performance, (2) a three-item measure thatexcluded the three innovativeness items, and (3) a three-itemmeasure that included only the innovativeness items. Thepattern of results was unchanged irrespective of the perfor-mance measure used.

Self monitoring was measured with the 18-item true-falseversion of the Self-Monitoring Scale (Snyder and Gangestad,1986). Items include "I would probably make a good actor,"and "In different situations and with different people, I often

act like very different persons." The seif-monitoring score,used as a continuous variable, indicates the probability thatan individual is a high or low self-monitor (Gangestad andSnyder, 1985). The shortened 18-item scale is both more reliable and more factorially pure than the original 25-item mea-sure (described in Snyder, 1974) with which it correlates at a.93 level (Snyder and Gangestad, 1986). In the presentresearch, Cronbach's (1951) alpha for the 18-item scale was.80.

The validity of the self-monitoring scale has been actively dis

cussed (see Snyder and Gangestad, 1986; Kilduff, 1992). Arecent comprehensive review pointed out that the most per-suasive evidence for the scale's predictive and constructvalidity consists of the several hundred studies of behavioraland attitudinal differences between high and low self-moni-tors consistent with seif-monitoring theory and detected bymeans of the Self-Monitoring Scale (Gangestad and Snyder,2000). With respect to discriminant validity, the Self-Monitoing Scale reliably predicts a range of criterion behaviors thatseem ingly similar scales, such as need for approval, locus ofcontrol, and field dependence, do not predict (Snyder, 1979).

Support for the stability of seif-monitoring comes from evi-dence that the latent causal variable corresponding to self-monitoring has a biological basis (Dworkin, 1977; Gangestad1984; Gangestad and Snyder, 1985): Monozygotic (MZ) twinare nearly always concordant on the iatent factor, whereasdizygotic twins are "concordant at better than a chance rate,but at a rate substantially iess than MZ twins" (Snyder andGangestad, 1986: 128). Additional support for the temporaistability of the self-monitoring scale comes from test-reteststudies over periods from one month to 3.5 months (summa-rized in Snyder, 1987: 17). Self-monitoring orientation can beunderstood as a distinctive aspect of each individual's person-ality, Accumulating evidence "suggests that seif-monitoring


14/27

ing individuals, by virtue of their control over resources andtheir decision-making authority, may be better positioned toemerge as central actors in social networks e.g., Lincoln andMiller, 1979; Ibarra, 1992). There were three levels of hierar-chy in the company. From company records, we coded rankas 0 for non-supervisors, 1 for supervisors, and 2 for top

management team members. enure The length of time a person has been with the com-pany is also likely to affect the pattern of participation insocial networks. For example, individuals who have beenwith the company longer may be more likely to occupy cen-tral positions in social networks. Using company records,tenure was coded as the number of months an individual hadbeen employed by the company.

Sex. We controlled for sex in each of the regression modelsbecause of its possible impact on network configuration Brass, 1985; Ibarra, 1993) and performance evaluation Burt,1992). Sex was coded as 0 for women and 1 for men.

Analysis

Our approach to testing the mediation, moderation, and addi-tive models follow s standard statistical procedures detailedin Baron and Kenny, 1986), We controlled for rank, tenure,and sex in each test, To assess support for mediation, weconducted three statistical tests to see if any significant rela-tion between self-monitoring and performance was eliminat-ed or significantly reduced once network position was con-trolled for. First, we used OLS regression to examine therelationship between self-monitoring and performance. Sec-ond, we used MANOVA to examine whether self-monitoringsignificantly pred icted the four netw ork variables taken as aset. Finally, to evaluate support for the overall mediationmodel, we used hierarchical regression analysis to examinewhether the inclusion of the four network variables signifi-cantly affected the relationship between self-monitoring andperformance. If a significant relationship between self-moni-toring and performance is eliminated or significantly reducedas a result of controlling for the four network variables, thenthis would indicate support for mediation.

We used hierarchical regression analysis to test the interac-tion model. To correct for the multicollinearity that ariseswhen testing moderated relationships among continuousvariables, we centered self-monitoring and the centrality vari-ables before generating interaction term s Cohen and Cohen,1983; Aiken and West, 1991). Centering consists of subtract-ing the sample mean from each independent variable. The

adjusted variables each have a mean of zero, but their sam-ple distribution remains unchanged. We co mp uted four inter-action terms by multiplying the centered self monitoring


15/27

Networks and Self monitoring

Testing the additive model was straightforward: self-monitoring and the four network variables were included simultane-ously as independent variables. If self-monitoring and thecentrality variables were significantly related to performancethen the additive model would be supported.

Size and betweenness centrality coUinearity. Despite the

clear conceptual distinction between the size of the individ-ual's network and the extent to which the individual's net-work links otherwise disconnected employees, size andbetweenness centrality are often highly correlated (BonacichOliver, and Snijders, 1998: 135). Popular individuals tend tohave high-betweenness centrality scores. Based on our theo-retical arguments, we were interested in examining howbetweenness centrality relates to dependent variables whilecontrolling for network size.

CoUinearity between variables such as size and betweenness

centrality tends to Inflate the standard errors of their regres-sion coefficients, making it more difficult to obtain significavalues, but the inflation of standard errors does not affect thevalidity of any significant results that are found. As oneregression expert explained, a significant value for the betacoefficient in a regression is just as conclusive w hencoUinearity is present as whe n it is absent (Darlington, 1990130).

To check on the severity of the multicollinearity between sizand betweenness centrality we examined the conditioning

index and variance proportions associated with each indepen-dent and control variable (see Belsley, Kuh, and Welsch,1980, for a discussion). According to Tabachnik and Fidell(1996: 86-87), a conditioning index greater than 30 and atleast two variance proportions greater than .50 indicates serious multicollinearity. None of our independent variables vio-lated this criterion; multicollinearity thus posed no seriousthreats to the validity of our analyses.

RESULTS

Table presents means, standard deviations, and zero-ordercorrelations among the variables. The typical employee hadbeen with the firm for 54 months. Men made up 62 percentof the sample. Individuals who were higher in rank, self-montoring, and betweenness centrality tended to have higher jobperformance ratings in these univariate tests. The density ofthe w ork flow network, as measured by the average cell valuin the 102 x 102 binary workflow matrix, was .34. The frienship network was considerably sparser, with a mean densityof .04.

The Mediation Model

According to the mediation model, the success of high self-monitors in outperforming low self monitors is due to the


16/27

Table 1

Means, Standard Deviations, and Correlations*

Variable Mean S.D.

1. Rank2. Tenure (months)3. Sex4. Self-monitoringWorkflow network5. Betweenness cenlrality6. SizeFriendship network7. Betweenness centrality8. Size9. Performance

0.3153.95

0.627,12

63.9049.27

146.637.24

20.25

0.6139.03

0.493.93

77.1919.50

243.615.505.08

.16

.14

.14

. 1 8

. 2 4 -

-.07.03. 36

-.06.07

04.04

. 3 3 * ^

.36**- . 2 6 -

.12

.10

.15

.01

.01

.02

.12

. 2 4 -

. 1 8 -

.04

. 2 3 -

. 8 7

.07

.13

. 2 6 -

.14

. 2 1 -

.17.80.04 -.10

p < 10; - p < .05; - p < . 0 1 ; p < . 0 0 1 .N = 93, except performance [N = 92}

tors. Controlling for rank, tenure, and sex. self-monitoring sig-nificantly predicted performance (p = 0 . 21 . p < .05), explain-ing an additional 4 percent of the variance over the baselinemodel .

Although high self-monitors may achieve higher job perfor-mance than low s elf-mo nitors, we still need to know if theyalso tend to occupy structurally advantageous positions insocial networks. The MANOVA results presented in the lastthree columns of table 3 show that controlling for rank,t nur and sex, self-monitoring significantly predicted thefour network variables taken as a set (F = 3.40, p < .05).explaining an additional 14 percent of the variance over thebaseline model . Table 3 also shows that higher self-monitor-ing scores predicted both higher betweenness centrality inthe friendship network and larger size in the workflow net-

Table 2

Standardized Regression Coefficients from Analyses Predicting Performance N = 92)

Model

Independent variable 1 2 3^ 4 5

Rank . 4 ( r " .3 8 .4 2 4 0 4 0 Tenure -.31 ^ -.32 - . 3 6 - . 3 7 - . 3 9 Sex -.06 -.08 -.07 -.08 -.09Setf-monitoring (SM) . 2 1 - . 19 - . 20 -Workflow network

Betweenness centrality .53* .59* .67*Size - . 3 7 - - . 4 7 - -.51

Friendship networkBetweenness central Ey , 4 1 - * .32 .28*Size - . 2 9 - -.22* -.19

SM X Workflow betweenness -.11

SM X Size of workflow network .11SM X Friendship betweenness -11SM X Size of friendship network --09


17/27


Table 3

Standardized Regression and MANOVA Coefficients from Analyses[N = 93)

Variable

RankTsnureSexSelf-monitoringModel F

Adjusted ^

Friendship Network

Betweennesscentrality

-.15. 3 4 .03. 1 7

3.92.15.11

p .10 ; - p < .05 ; p < , 01 ; p

Date post:	03-Jun-2018
Category:	Documents
Upload:	sibel-ceaus
View:	220 times
Download:	0 times

Mehra Kilduf Brass 2001 Self Monitors

Documents