Job Satisfaction and Teamwork: The Role of Supervisor Support Author(s): Mark A. Griffin, Malcolm G. Patterson, Michael A.

West Source: Journal of Organizational Behavior, Vol. 22, No. 5 (Aug., 2001), pp. 537-550 Published by: John Wiley & Sons Stable URL: http://www.jstor.org/stable/3649557 Accessed: 02/02/2010 23:31
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=jwiley. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

John Wiley & Sons is collaborating with JSTOR to digitize, preserve and extend access to Journal of Organizational Behavior.

http://www.jstor.org

Journal of Organizational Behavior J. Organiz. Behav. 22, 537-550 (2001) DOI: 10.1002/job.101

Job satisfaction and teamwork: the role of supervisor support

MARK A. GRIFFIN1*, MALCOLM G. PATTERSON2 AND MICHAEL A. WEST3
1QueenslandUniversityof Technology,Brisbane, Australia 2The Universityof Sheffield,Sheffield, U.K. 3The Universityof Aston, Aston, U.K.

Summary

The link between teamworkandjob satisfactionwas investigatedin a sample of 48 manufacturing companies comprising 4708 employees. Two separate research questions were addressed.First, it was proposed that supervisor supportwould be a weaker source of job satisfaction in companies with higher levels of teamworking.Multilevel analysis indicated that the extent of teamwork at the company level of analysis moderated the relationship between individualperceptionsof supervisorsupportandjob satisfaction.Second, it was proposed thatthe extent of teamworkwould be positively relatedto perceptionsof job autonomy but negatively related to perceptionsof supervisorsupport.Further,it was proposedthat the link between teamworkand job autonomy would be explained by job enrichmentpractices associated with teamwork.Analyses of aggregatedcompany data supportedthese propositions andprovidedevidence for a complex mediationalpathbetween teamworkandjob satisfaction. Implicationsfor implementingteamworkin organizationsare discussed. Copyright ? 2001 John Wiley & Sons, Ltd.

Introduction
Teamwork typically involves groups of interdependent employees who work cooperatively to achieve group outcomes (Parker and Wall, 1998). Effective team implementation can enhance the motivational properties of work and increase job satisfaction. However, the job satisfaction of team members is determined by multiple factors such as the composition of the team, group processes within the team, and the nature of the work itself (Campion et al., 1993; Gladstein, 1984). Because these factors operate in combination, there is no simple process through which teamwork influences job satisfaction. Despite the potential advantages of teamwork, the introduction of teams sometimes fails to result in expected outcomes for individuals and organizations (Hackman, 1990). This paper explores ways in which changes in leadership roles influence overall job satisfaction in teams. Within the broad field of leadership, we focus on employee experiences of supervisor support: the extent to which supervisors provide encouragement and support to employees within their work
* to: MarkA. Griffin,School of Management,QueenslandUniversityof Technology,Brisbane,4001, Australia. Correspondence E-mail: m.griffin@qut.edu.au

Copyright? 2001 John Wiley & Sons, Ltd.

Received 5 June 1999 Accepted 6 November 2000

Published online 26 June 2001

538

M. A. GRIFFINETAL.

groups. The supportand considerationof supervisorsis a strong determinantof job satisfactionin a aboutthe role of leaderwide varietyof work settings(Yukl, 1989). Theoretically,betterunderstanding in teams some research inconsistent may help explain findingsconcerningthe impact of ship changes teamworkon satisfaction.On a practicallevel, it is importantto identify factors that will facilitate the effective implementationof teams in the work place. We exploredthe relationshipbetween teams andsatisfactionusing datafrom a large researchproject in the U.K. that investigatedmanagementpractices and employee attitudesin manufacturing companies. Data from the projectwere used to assess researchquestionsaboutjob satisfactionat two levels of analysis. First, we focused on the individuallevel of analysis to investigate the importanceof supervisor support for job satisfaction when teams are introducedin companies. This level of analysis allowed us to assess the degree to which the link between supervisor support and job satisfaction was influencedby the extent of teamworkin each company.Second, we focused on the companylevel of analysis to investigate differences in average satisfaction levels across companies. This analysis enabled us to differentiatethe role of teamworkfrom more generaljob enrichmentpractices and to investigate how these practices were linked, via supervisorsupport,to job satisfaction.Analyses at the different levels draw on the strengthsof a multi-organizationsample in which informationwas obtained from individualemployees, managerinterviews, and company records.

Individual Level: The Changing Impact of Supervisor Support

role in structuring the work environmentand providinginformationand Supervisorsplay an important feedback to employees. As a consequence, supervisorbehaviorshave an impact on the affective reactions of team members (Durhamet al., 1997). However,introducingteams can result in a significant et al., 1996). Parkerand Wall change to the role of supervisorswithin organizations(Tannenbaum (1998) identify a numberof options for leadershiproles in teams. These options range from the complete elimination of supervisorypositions to the retentionof supervisorypositions but with redefined role requirementssuch as facilitation. All of these options change the role of supervisorsin teams so that supervisionis a less importantsource of supportand direction (Kerret al., 1986). Our first goal was to investigate the impact of supervisorsupporton employee satisfactionunder different levels of team implementation.Little researchhas tested whether the impact of leadership changes when teams are introduced.One possible consequence of the changing role of supervision when teams are introducedis thatthe traditionalsupportprovidedby supervisorsbecomes less importantfor employee satisfaction.Therefore,there shouldbe a weakerlink between the perceivedsupportiveness of supervisorsand the satisfactionlevels of employees in companies where teams have been introduced.The companies in the present study differed in the degree to which they had introduced teamwork. Some companies had made no change toward the use of teams, some had implemented teamwork across the whole organization, and others had introducedminimal elements of teamworking. Although leadershipmay have less influence on satisfactionfor employees working in teams, this does not mean that a supervisor'ssupporthas no impact on satisfaction. Leadershipis consistently recognized as importantfor the initiation and ongoing development of teams (Bass, 1997; Manz and Sims, 1987, Tjosvold, 1995) and is often included as an importantdeterminantin models of team satisfaction(e.g., Campionet al., 1993; Cohen et al., 1996; Gladstein, 1984). In traditionalwork structures,supervisors have long been recognized to play an importantpart in developing roles and expectations of employees (Graen and Scandura, 1987). This function is also importantwhen
Copyright? 2001 John Wiley & Sons, Ltd. J. Organiz.Behav. 22, 537-550 (2001)

AND TEAMS JOB SATISFACTION

539

teamworkis introducedbecause supervisorscan play a key role in modelling teamworkand setting the groundrules for team membersto engage in team processes (McIntyreand Salas, 1995). Immediate supervisorsalso provide salient informationabout the supportof the broaderorganizationfor change and their behaviors are likely to be interpretedas representativeof wider organizationalprocesses (Kozlowski and Doherty, 1989). Therefore,althoughwe proposethat supervisorsupportwill be a less importantsource of satisfactionwhen teams are introduced,we also propose that supervisorsupport will continue to exert a positive influence on employee satisfaction. Based on the above discussion, hypotheses were developed to predict the overall impact of supervisor supporton job satisfaction,variationin the impact of supervisorsupporton satisfactionacross companies,and a link between the extent of teamworkand the importanceof supervisorsupport.First, we proposedthat the overall link betweenjob supervisorsupportandjob satisfactionwill be positive regardlessof the level of teamworkin an organization. Hypothesis 1: Supervisor support will have a positive impact on job satisfaction across all companies. Next we proposedthattherewould be systematicvariationin the link between perceptionsof supervisor supportand satisfactionacross the range of companies in the study. It is importantto establish this variationbecause it demonstratesthat the link between supervisorsupportand satisfactionis not the same within differentcompanies. Hypothesis2: The relationshipbetween perceptionsof supervisorsupportandjob satisfactionwill vary systematically across companies. If hypothesis 2 is supported,it allows the possibility that variationin the link between supervisor supportand satisfactioncan be explained by company level differences. Our key hypothesis was that these differences could be explained by the extent of teamworkin each company.Therefore,the following hypothesis was tested. Hypothesis 3: Supervisorsupportwill be more strongly related to job satisfaction in companies with higher levels of teamworkcomparedto companies with lower levels of teamwork. The study also assessed otherorganizationalcharacteristics that might influencejob satisfaction.In particular,the performanceof the company is likely to influence the level of job satisfaction of employees in the company (Ryan et al., 1996). Companieswith higher levels of productivityand performance are likely to have more resources, greater security, and provide more opportunitiesfor employees. The directionof the effect between organizational performanceand employee satisfaction has received little researchattention(Ostroff,1992). The currentstudyused productivityas a company level controlmeasureto accountfor otherfactors not associated with teamwork.Similarly,the size of the organizationalunit may have an impact on individualjob satisfactionbecause smaller establishments tend to be asociated with more satisfied workers (LaRocco, 1985; Steers and Rhodes, 1987). Analysis by Clark(1996) suggested thatgreaterprovision of intrinsicrewards(e.g., relationsat work) from smallerorganizationsresults in significantlyraised satisfactionin those workerswho value such characteristics. Therefore,organizationalsize was also included as a control variable.The following hypotheses tested the role of these two measures at the company level of analysis. Hypothesis 4: Companies with higher profit levels will report higher overall levels of job satisfaction. Hypothesis5: Companieswith more employees will reportlower overall levels of job satisfaction. ? 2001 JohnWiley& Sons,Ltd. Copyright J. Organiz. Behav. 22, 537-550 (2001)

540

M. A. GRIFFINETAL.

Figure 1. Hypothesizedcross-level influences on job satisfaction

Figure 1 summarizesthe hypotheticalrelationshipsthat were investigated.The figuredifferentiates measuresobtainedat the individuallevel of analysis from those obtainedat the companylevel of analysis. At the individuallevel, perceptionsof supervisorsupportandjob satisfactionwere assessed. At the organizationallevel, the extent of teamwork,profitability,and company size were assessed.

Company Level: The Mediational Role of Supervisor Support and Job Autonomy
In additionto changing the role of supervisors,the introductionof teams is typically accompaniedby changes to the design of work roles for employees. We focused on job enrichmentpracticesas a common practicecarriedout by organizationsaiming to redesign the work place. These practicesinclude increasing the responsibilities of employees, the variety of tasks performed, and the flexibility to implementtasks. Researchin the area of job enrichmenthas consistently demonstrated positive links to employee satisfaction (Hackman and Oldham, 1980; Neuman et al., 1989). A key outcome for employees from job enrichmentpractices is the experience of job autonomy: the extent to which employees experience a sense of choice and discretion in their work. There is also a long line of researchsupportingthe positive impact of job autonomyon job satisfaction(Parkerand Wall, 1998). Job enrichmentpractices,such as increasingjob variety,can resultin higherjob satisfactionbecause of higher levels of perceived autonomy.The introductionof teamworkis also often undertakenwith the goal of creating an enrichedwork environmentin which employees can experience greaterautonomy (Wall et al., 1986). Evidence suggests that teams with higher levels of autonomyare more effective (Beekun, 1989). However,job enrichmentpractices are not necessarily implementedthrougha teamwork structure.For example, job variety may be increased through a job rotation programme or through a team structurein which team members share all tasks. In both cases we expect that the enrichmentpractice of increasedjob variety will result in higher levels of job satisfaction.However, where job enrichmentis part of a teamworkwe expect less direct supportfor employees from supervisors.In this way, supervisorsupportcan be lower when teams are introduced.
Copyright? 2001 John Wiley & Sons, Ltd. J. Organiz.Behav. 22, 537-550 (2001)

ANDTEAMS JOBSATISFACTION

541

Measuresobtained fromcompany
Job PathC

Measuresobtainedfrom individuals
Job

Enrichment

thPathPa .fpa+ PathD Extent of / ' Teamwork B Path B |I~ ~ Control Measures

Eath

Autonomy
PathSp

Job Job
Satisfaction

Supervisory Support

Figure 2. Hypothesizedrelationshipsamong organizationallevel constructs

We hypothesized that job enrichment associated with the introduction of teams results in greaterautonomyfor employees but that team structuresthemselves reduce levels of supervisorsupport. In this way, teamworkcan indirectlyinfluencejob satisfactionin both a positive and a negative direction.The proposednegative impact of teams on supervisorsupportis implicit in the literatureon teams but has not been tested directly. Supportfor this negative impact would provide one explanation for the difficulties associated with the introduction of teams and suggest an important mechanism for managing the transition from more traditional work structures to teamwork. A negative impact of teamworkon supervisorsupportwould highlight the need to ensure supportfor the functional role of leaders particularlyduring transition.Overall, if the introductionof teams is not associated with effective adaptationof supervisoryroles, there may be negative outcomes for teams (Komakiet al., 1989). Figure2 depicts the proposedmediationalrole played by job autonomyand supervisorsupportat the company level of analysis. This level of analysis allowed us to investigate the mediationalprocesses throughwhich teamworkand leadershipperceptionsinfluence job satisfaction. The model does not propose that satisfaction is a collective constructbut provides an opportunityto investigate factors associated with average differences across companies. The figure distinguishes between the extent of teamwork in an organizationand the degree of job enrichmentin the workplace. Although the two constructs are distinct, teamwork and job enrichmentwere expected to be correlated(Path A) because the introductionof teams is typically associatedwith changes in job design (Parkerand Wall, of teams was expected to have a directnegativeimpact on levels of supervisor 1998). The introduction support(PathB). That is, the introductionof teams itself can reduce the level of supervisorsupportin organizations.Job enrichmentwas expected to lead to higher levels of job autonomy(Path C). There was expectedto be no directeffect of the extent of teamworkon job autonomy(PathD) as this relationship is explainedby the correlationwithjob enrichment.Job enrichmentthatprovidesgreaterflexibility and control was expected to have no direct influence on the levels of supervisorsupportwithin teams (PathE). Job autonomyhas been found to mediatethe impactof job enrichmenton team outcomes (Janzet al., 1997). Therefore,both supervisorsupportandjob autonomywere expected to have a positive impact on job satisfaction(Path F and Path G, respectively). In summary, we proposedthattypical team interventionsmay reducethe level of supervisorsupport in the work place by focusing on increased flexibility and autonomy for employees. However, if the
Copyright? 2001 John Wiley & Sons, Ltd. J. Organiz.Behav. 22, 537-550 (2001)

542

M. A. GRIFFINETAL.

ContextualSidebar
The companies The companieswere spreadthroughout the United Kingdon,and were drawnfrom the manufacture of metal goods and mechanicalengineeringsector,plastics and rubbersector, and a thirdmiscellaneous category.They were predominantlysingle site and averagedapproximately260 employees. Almost all the firmsreportedundergoingsignificantchange in structureand work design. Over 80 per cent of companies reported greater decentralization.However, when senior managers were asked to indicate the lowest hierarchicallevel where employees had the authorityto make selected decisions, responses revealed that supervisorsand operatorsmostly had limited opportunitiesfor decision making. Workers' views of the workplace Organizationalclimate surveys conducted across the companies reinforced the perception that changes in HRM and empowerment were implemented with limited effectiveness. Employees Presreportedlow levels of participation,lack of open communication,and little decentralization. sure on employees to produceand formalization(the use of rules and formalproceduresto control activities) dominatedemployees' perceptionsof their organizations. Time The study was conductedbetween 1994 and 1995.

positive impact of supervisorysupportis reduced,then the overall change in satisfactionmay be less than expected if the change is not compensatedfor by other changes in the workplace.

Method
Sample and procedure
The companies in this study were 48 manufacturingcompanies in the U.K. that were involved in a largerprojectexaminingorganizational design, managementpractices,employee attitudes,innovation andeconomic performance. The companiesrangedin size from 60 to 1929 employees, with an average of 246 employees (SD = 318.8). Teams had been introducedto varying degrees across the range of companies. The companies were predominantlysingle site and single productoperations. Measuresat the individuallevel of analysis were obtainedthrougha survey of all staff in each company.Data were collected on employee job satisfaction,and employee perceptionsof supervisorysupport and work autonomy. In most companies 100 per cent of employees were surveyed but in companies with more than 500 employees (n = 3) a randomsample was taken. Response rates within companies ranged from 5-100 per cent with a mean response rate of 52.5 per cent (SD = 26.4 per cent). The questionnairesurveys were either distributedto employees by the organizationor handed to employees on-site by a memberof the researchteam. Listwise deletion of missing cases resultedin a final sample for analyses of 4708 in 48 companies. Individualmeasuresused in the second analysis were aggregatedto the company level of analysis.
Copyright ? 2001 John Wiley & Sons, Ltd. J. Organiz.Behav. 22, 537-550 (2001)

JOB SATISFACTION AND TEAMS

543

Measuresof teamworkingandjob enrichmentwere collected at the company level throughon-site, interviews with senior managersand directors.The interview information in-depth, semi-structured formed the basis of researcherratings of teamworkingand job enrichment.We did not rely on managers' ratingsbecause theirresponseswere not similarlyanchoredacross companies.For example, an inexperiencedmanagerin one companymight rate a ratherlimited implementationof job enrichment as disproportionally substantial.The field-researchers' had contact with managersand experience of work conditions across all companies in the study.

Individual level measures

Job satisfaction was measuredby Warret al. (1979) 15-item measure (alpha= 0.92). The scale was in previous researchon blue-collar employees (Cook et al., 1981; Cook developed and standardized and Wall, 1980). The responseformatrangedfrom extremelydissatisfied(1) to extremelysatisfied (7). cliSupervisorysupportwas measuredby five items (alpha= 0.88) from an index of organizational mate developed for use with blue-collarworkers(Lawthomet al., 1997). The response formatranged from definitely false (1) to definitelytrue (4). An example item was 'Supervisorshere are friendly and to easy approach'. Workautonomywas measuredby four items (alpha= 0.68) from the same index of organizational climate used to assess supervisorysupport(Lawthomet al., 1997). The response formatrangedfrom definitely false (1) to definitelytrue (4). An example item was 'Managementlet people make theirown decisions most of the time'.

Company level measures

Teamwork The interviewquestionscovered the degree to which teams, ratherthanindividuals,had responsibility for work, the structureof the teams, how they were managed and how targets were set. Using this information, researchersrated the degree to which teamworking had been implemented. Scoring was on a 4th-point scale from 'not at all' to 'extensively'. Job enrichment The interview covered the following aspects of job design: the extent of job rotation;the extent of multiskillingof shopflooremployees; the cycle time in the typicaljob; the extent to which the management of materialsand work in progresswas delegated to shopfloorstaff; the amountof responsibility operatorshad for dealing the technology, including minor breakdowns, routine maintenance and machine set-up; and the discretionthey had for deciding on when to take breaksand for the methods used to complete theirwork.Using this information,complementedby documentary(e.g., job descriptions) and observationalevidence (e.g., shopfloor behaviour and comments), researchersrated the degree of job enrichmenton the following dimensions:the extent of skill flexibility of shopfloorworkers, the extent of job variety, and the degree of job responsibility.Scoring was on an appropriately worded 7-point scale running, for example, from very low (1) to very high (7). Cronbach'salpha for job enrichmentwas 0.84. Control measures Controlmeasuresincludedwere firmsize and companyproductivity. Firm size was representedby the logarithm(to compensatefor skewness) of total employment.Following priorwork (Pritchard,1992, Huselid, 1995), productivitywas measuredby the logarithmof sales per employee.
Copyright? 2001 John Wiley & Sons, Ltd.

J. Organiz. Behav. 22, 537-550 (2001)

544

M. A. GRIFFIN ETAL.

Table 1. Correlationsamong all measurescalculated at the individuallevel (n = 4708) and at the company level (n = 48) 1 1. Teamwork 2. Size 3. Profit 4. Job enrichment 5. Supervisorsupport 6. Satisfaction 7. Autonomy 1.00 -0.171 0.11t 0.55* - 0.101 -0.05t 0.08t 2 0.02 1.00 0.37t -0.21t - 0.02 0.07t 0.02 3 0.01 0.11 1.00 0.21t -0.07t O.10t 0.04t 4 0.50t -0.03 0.25* 1.00 -0.01 0.04t 0.15t 5 -0.25* 0.09 0.07 0.01 1.00 0.56t 0.22t 6 -0.18 0.10 0.38t 0.14 0.34t 1.00 0.32T 7 0.26* 0.07 0.20 0.50t 0.01 0.34t 1.00

*p< 0.05;tp< 0.01;tp< 0.001. Note: Individual level correlations below the diagonal,aggregate level correlation above the diagonal.Individual level correlations in the company level measures and company to each individual level by disaggregating computed company correlations individual measures within by aggregating computed companies.

Correlationsamong the variables at both the individual and the aggregate level are reportedin
Table 1.

Analysis
Hypothesesinvolving satisfactionat the individuallevel requireda cross-level analysis in which organizational characteristics moderate relationships among individual perceptions (Rousseau, 1985; Hofmann et al., 2000). The hypotheses were assessed using hierarchicallinear models (HLM, Bryk and Raudenbush,1992). The HLMan procedureis an appropriate procedurefor linking company level
measures and individual measures when the dependent variable is assessed at the individual level of

analysis. The HLM procedureis a two stage strategythat, in the first stage, estimates within-company parameters separatelyfor each of the companiesusing the individuallevel variables.Specifically,individualjob satisfactionwas regressedon supervisorsupportwithin each of the 48 companies.This anaacross all companies.The averageslope lysis providesestimates of the interceptand slope parameters of satisfactionregressedon supervisorsupportestimatesthe overall link between these two individual measures (Hypothesis 1). The variancein the interceptterm representsmean differences in job satisfaction across companies after controllingfor individualperceptionsof supervisorsupport.Across all companies, the variance in the slope of satisfaction regressed on supervisor supportestimates the degree to which the relationshipbetween supervisorsupportand satisfactionvaries across companies (Hypothesis2). The second stage of the analysis investigatescompanylevel predictorsof the variance in the interceptand slope. If there is significantvariationin the link between supervisorsupportand satisfaction,the second stage of the analysis tests whethercompany level measureof teamworkpredicts variation in the relationship among the individual variables (Hypothesis 3). The second stage of

the analysis also tests whethervariationin the interceptis predictedby the companylevel measuresof profitabilityand company size (Hypotheses 4 and 5). Structural equationmodeling (SEM) was used to test the mediationalmodel incorporating job satisfaction at the companylevel of analysis. LISRELVIII (Joreskogand Sorbom, 1993) was used to estimate the paths depicted in Figure 2 and to estimate the overall fit of the hypothesizedmodel. Because of the relatively small sample size at this level, scale scores were used to assess each construct(Bentler and Chou, 1988). For measures based on aggregatedindividual scores, the measurementerrorwas estimatedas being equal to the varianceof the scale multipliedby one minus the reliabilityof the scale (Andersonand Gerbing,1988). The measuresof teamworkandjob enrichmentwere assumedto have a reliability of 0.9 and company profit and size were assumed to have no measurementerror.
Copyright ? 2001 John Wiley & Sons, Ltd. J. Organiz.Behav. 22, 537-550 (2001)

JOB SATISFACTION AND TEAMS

545

Results
The first step in the HLM analysis estimated the impact of supervisory support on satisfaction within companies. Table 2 reports the results for this analysis. The pooled value for the effect of supervisory support was significantly positive, indicating that supervisory support had an overall positive impact on satisfaction across all companies (/ )= 0.91, p < 0.001). The results of the HLM analysis also estimated variance of 0.15 in the parameterestimates around the pooled estimate of 0.91 indicatingthatthe impactof supervisorysupportwas positive across all 48 companies,supporting Hypothesis 1. Table 2 also shows the variancein this pooled estimate with a chi-squaretest of the statisipotogether tical significanceof the variation.The interceptvalue representsthe mean level of satisfactionacross the 48 companiesaftercontrollingfor supervisorysupport(Hofmannet al., 2000). The variationin the interceptterm describes the differences across companies in mean levels of satisfaction. The HLM results show that there was significant variation in mean satisfaction across organizations (2(47) = 692.14, p < 0.001) and significantvariationin the slope of satisfactionwhen regressed on supervisorysupportperceptionsin each organization(2(47) = 390.10, p < 0.001). The latter result supportsHypothesis2 which proposedtherewould be significantvariationin the relationshipbetween supervisorysupportand job satisfactionacross companies. The next step in the HLM analysis was to introducecompanylevel predictorsof the variationacross companies.The extent of teamwork,company size, and companyprofitabilitywere enteredas predictors of variationin mean levels of satisfaction (intercept),and variationin the slope of satisfaction regressedon leadership.Table 3 summarizesthe results of this stage of the HLM analysis. In support of Hypothesis 3, the extent of teamworkwas a significantpredictorof the variationin the slope of satisfactionregressedon leadership(0= - 0.147, p < 0.05). The directionof the result indicatedthat the impact of leadershipon job satisfactionwas weaker when organizationsreportedhigher levels of teamwork. The results in Table 3 also show that profitabilitywas positively associated with average levels of satisfactionacross companies (0 = 0.238, p < 0.01). SupportingHypothesis4, Companieswith higher profitlevels reportedhigherlevels of job satisfaction.Companysize was not relatedto variationin any of the within-companymeasures.Therefore,Hypothesis 5 was not supported. The results of the HLM analysis supportedthe propositionthat supervisorysupportis a less important source of individualjob satisfaction when there was greater use of teams within a company. Although supervisorysupporthad less impact on satisfaction,the results also indicated that supervisory supportremaineda significantinfluenceon job satisfaction.Therefore,when teams were used to a greaterextent, leadershipbecame less importantbut not unimportant.

Table 2. Results of HLManalysis individual levelmeasures of including andsupervisor job satisfaction support Predictor Pooled T-value parameter estimate 84.47* 14.23* Parameter Chi-square variance value 0.12 0.15 692.14* 390.10*

4.45 Intercept Supervisor support 0.91

*p < 0.001.

Copyright? 2001 John Wiley & Sons, Ltd.

J. Organiz.Behav. 22, 537-550 (2001)

546

ETAL. M. A. GRIFFIN Table3. Results of HLM analyses incorporating companylevel predictors Predictor Predictionof intercept Teamwork Size Profitability Predictionof relationshipbetween supervisorysupportand satisfaction Teamwork Size Profitability *p< 0.05;tp< 0.01. Parameter estimate - 0.059 -0.003 0.238 - 0.147 - 0.067 - 0.080 T-value

-1.11 - 0.04 2.41t - 2.14* 0.79 - 0.68

The next part of the study explored implications of the mediationalprocesses linking teamwork,
supervisory support, and job satisfaction. The analysis tested the hypothetical model depicted in Figure 2. The SEM model provided a very good fit to the data (X2(3) = 3.22, p > 0.05, GFI = 0.98).

The fit of the model suggested that additional paths were not necessary to explain the covariation among the measures.The final estimates for each path are presentedin Figure 2. The sign and statistical significanceof the paths supportedthe hypotheticalmodel. The extent of teamworkin organizations was negatively related to perceptions of supervisory support (3 = - 0.42, p < 0.05) and was not
related to perceptions of job autonomy (o = 0.04, n.s.). Job enrichment was positively related to job

autonomy(/ = 0.60, p < 0.01) but was not significantlyrelatedto perceptionsof supervisorysupport (p = 0.23, n.s.). Both job autonomyand supervisorysupportwere significantpredictorsof job satisfaction (/ = 0.31, p < 0.01 and 3= 0.36, p < 0.01 respectively). The control variableof company profit was entered in the model as a predictorof all other measures. The results indicated that profit was significantlyand directly related only to job satisfaction (P = 0.30, p < 0.05). The results of the SEM analysis supportthe propositionthat teamworkcan have a complex impact on job satisfactionbecause of the differentways that teamworkrelates to job autonomyand supervisory support.While both autonomyand supervisorysupportwere important positive influenceson job satisfaction,the positive association of teamworkon job autonomywas explained by job enrichment that accompaniedteamworkand could not be attributed to teamworkitself.

Figure 3. Final path estimates from SEM analysis

Copyright ? 2001 John Wiley & Sons, Ltd.

J. Organiz.Behav. 22, 537-550 (2001)

AND TEAMS JOB SATISFACTION

547

Discussion
of teamworkhas potentialbenefitsfor employee outcomes in organizationsand is The implementation likely to continue as a major element of work redesign (Lawler et al., 1992; Parkerand Wall, 1998; Tjosvold, 1995). However,greateruse of teams is associatedwith changes to the functionof leadership in organizationsthat can also influence employee satisfaction.The present study explored the role of supervisorysupportin explaining the link between teamworkand satisfaction. In the first part of the study, greateruse of teams resulted in a weaker link between individuals' perceptions of supervisory support and their levels of job satisfaction. The implementation of teams typically involves changes to supervision roles but little empirical research has assessed the way leadership functions when teams are introduced.The study, therefore, provides evidence for the implicit assumption that supervisory support has less impact on employees when teams are introduced. Even though supervisory support was less important in companies where there was greater use of teams, this support was still positively related to satisfaction. Therefore, it is not appropriateto conclude that supervisory supportwas unimportantwhen there was greater use of teams. The second partof the study exploredfurtherimplicationsfor the role of supervisorsupportby testing mediationalprocesses linking teamworkto job satisfaction.The extent of teamworkin a company had a negative impact on overall levels of supervisorysupport.Companiesthatreportedhigher use of teamworkalso had employees who reportedlower levels of supervisorysupport.At the same time, supervisory supportdisplayed a strong relationshipwith job satisfaction. Therefore, the impact of teamworkitself can have a negative impact on overall job satisfactionbecause of the central role of supervisorysupport. The link between teamwork and satisfaction was partly explained by job enrichment associated with teamwork.Job enrichmentpractices in the work place were associated with increased levels of employee perceptions of job autonomy. Increased autonomy was, in turn, associated with higher levels of job satisfaction. However, on its own, teamworkwas associated with a reductionin supervisory support.The result suggests that the positive benefits of job enrichmentassociated with teamwork may be partially offset by the reduction in levels of supervisory support experienced by employees. In this way, the introduction of teamwork can have mixed effects on employee job satisfaction. The extent of teamworkwas not uniquely related to higher levels of job autonomy.Rather,autonomy was explainedby the degree to which employee jobs were enrichedby increasedskill flexibility, the importanceof identifyingthe proresponsibility,andwork variety.This resultclearly demonstrates cess throughwhich the introductionof teamwork is expected to enhance employee outcomes. The introductionof teams without reference to specific job enrichmentpractices is unlikely to influence employee perceptions about the quality of their work environment.Furthermore,teamwork itself can be expected to decrease the degree to which supervisorscan be supportive. Some implicationscan also be derivedby the relationshipswhich were not observed in the aggregated mediationalmodel. Job enrichmentpracticesreportedby managerswere not uniquelyrelatedto averagelevels of supervisorysupportacross the organizations.It was noted above that leaders play a particularly importantrole when organizationsincrease their use of teams (Tjosvold, 1995). It is possible for changes in work structuresto be implemented in ways that capitalize on the potential for supervisorsand leaders to supportand develop new working roles in the organization.For example, increasedjob varietymay be implementedin an organizationwhere team leaders,supervisor,andmanagers encourage and supportskill development of employees. Furthermore,supervisorscan play a

Copyright? 2001 John Wiley & Sons, Ltd.

J. Organiz.Behav. 22, 537-550 (2001)

548

M. A. GRIFFINETAL.

significant part in enhancing the motivational characteristicsof the work environmentsuch as job autonomy.Effective leaders provide feedback aboutroles and tasks that increase employees' positive experience of autonomy. The present study was based on a large scale survey of employees across multiple organizations together with in-depth management interviews. The scope and multiple methods provide a strong basis for exploring the implications of teamworkfor job satisfaction. Some limitations of the study also raise cautions when interpretingthe results. First, the cross-sectional natureof the study limits the ability to draw causal inferences from the hypothetical model of relationships among the constructs. For example, the relative ordering of job satisfaction and supervisory supportcould not be tested in the currentsample. Second, despite the scale of the study,there was a relatively small number of organizationalunits for conductingthe aggregate-levelanalysis. Structural equationmodelling procedurestypically use larger samples than were available in the currentstudy. Nevertheless, relatively few parameterswere assessed and specific hypotheticalpaths were tested. Third,the extent of teamworkwas assessed by a single item which limits assessment of the measure's reliability. However, responses to the item were obtained as part of a comprehensive interview procedure with detailed explanation, and single-item measures of practices have been used in major empirical studies in the field of human resource management and corporate performance (e.g., Huselid, 1995, Ichniowski et al., 1997). Futureresearch should furtherinvestigate the processes throughwhich supervisorbehavior influences satisfaction in teams. The type of supervisor behaviors that are most supportive may be different in a team environment compared to an individual based task environment. The most effective supervisor behaviors may also depend on individual differences among team members within the team. Furtherunderstandingof how supervisorsprovide supportwithin teams will have practicalbenefits for designing and managing teamwork.Futureresearch should also investigate the characteristicsof job design and teamwork that enhance the role that supervisors play in a team environment as well as characteristicsof supervisors that lead to higher levels of supportiveness in teamwork. thatthe extent of teamworkreducesthe level of supervisorsupport Overall,the resultsdemonstrated in the organizationand, to some extent, the importanceof supervisorysupportfor individuals.However, the perceptionof supervisorysupportremaineda substantial predictorof job satisfactionfor individuals and for the organization as a whole. The results of the study suggest it is importantto understandjob redesign strategies that not only increase experiences such as autonomy, but also enhance the effectiveness of supervisorroles.

Author biographies
Mark A. Griffin is PrincipalResearchFellow at the School of Management,QueenslandUniversityof Technology, Australia. He received his PhD in Industrial/Organizational Psychology from the Pennslyvania State University. His primaryresearch interest concerns the link between workplace characteristicsand both individualwell-being and performance.His researchaddressesthe methodological and practicalimplicationsof questions such as the impact of organizationalclimate on individual affect, and the impact of individualperformanceon workplace safety outcomes. Malcolm Patterson is a Research Associate at the Institute of Work Psychology, University of Sheffield. He received his master's degree in occupational psychology from the University of Sheffield.His currentresearchexamines the relationshipsbetween humanresourcemanagementpractices, employee attitudesand behaviors, and firm performance. ? 2001 JohnWiley& Sons,Ltd. Copyright
J. Organiz.Behav. 22, 537-550 (2001)

ANDTEAMS JOBSATISFACTION

549

Michael West is Professor of Organizational Psychology at the University of Aston Business School.

He has been a memberof the CorporatePerformanceProgrammeof the Centrefor Economic Performance at the London School of Economics since 1991. He is a Fellow of the British Psychological Society, the American Psychological Association, the APA Society for Industrial / Organizational Psychology and the Royal Society for Arts, Manufactures and Commerce. His areas of research interest

are team and organizational innovationand effectiveness, in health service and privatesector settings.
His first degree was from the University of Wales in Psychology and Statistics and his PhD from the University of Wales on the Psychology of Meditation.

References
AndersonJC, GerbingDW. 1988. Structural equationmodeling in practice:a review and recommendedtwo-step approach.Psychological Bulletin 103: 411-423. Bass BM. 1997. Does the transactional-transformational leadership paradigm transcend organizational and nationalboundaries.AmericanPsychologist 52: 130-139. Beekun RI. 1989. Assessing the effectiveness of sociotechnical interventions:antidoteor fad. Human Relations 42: 877-897. Bentler PM, Chou CP. 1988. Practical issues in structuralmodeling. In CommonProblems/ProperSolutions: Avoiding Error in QuantitativeResearchLong JS. (ed.). Sage: Newbury Park,CA; 161-192. Bryk AS, RaudenbushSW. 1992. HierarchicalLinear Models. Sage: Newbury Park,CA. Campion MA, Medsker GJ, Higgs AC. 1993. Relations between work group characteristicsand effectiveness: implicationsfor designing effective work groups. Personnel Psychology 46: 823-850. ClarkAE. 1996. Job satisfactionin Britain.British Journal of IndustrialRelations 34: 189-217. Cook JD, HepworthSJ, Wall TD, WarrPB. 1981. The Experienceof Work: A Compendium of 249 Measuresand their Use. Academic Press: London. Cook J, Wall TD. 1980. New work attitudemeasuresof trust,organizational commitmentand personalneed nonfulfilment.Journal of OccupationalPsychology 53: 39-52. Cohen SG, Ledford GE, Spreitzer GM. 1996. A predictive model of self-managing work team effectiveness. HumanRelations 49: 643-678. DurhamCC, Knight D, Locke EA. 1997. Effects of leader role, team-set goal difficulty,efficacy, and tactics on team effectiveness. OrganizationalBehavior and HumanDecision Processes 72: 203-231. GladsteinDL. 1984. Groupsin context:a model of task groupeffectiveness.Administrative Science Quarterly29: 499-517. GraenG, Scandura TA. 1987. Towarda psychology of dyadic organizing.Researchin OrganizationalBehavior9: 175-208. HackmanJR. 1990. GroupsThat Work(and Those ThatDon't). Jossey-Bass: San Francisco. HackmanJR, OldhamGR. 1980. Work Redesign. Addison-Wesley:Reading, MA. Hofmann DA, Griffin MA, Gavin M. 2000. The application of hiearchical linear modeling to management research.In Multilevel Theory,Research, and Methods in Organizations.Klein K, Kozlowski S (eds). Jossey Bass: San Francisco. HuselidMA. 1995. The impactof humanresourcemanagementpracticeson turnover, and corporate productivity, financialperformance.Academyof ManagementJournal 38: 635-672. IchniowskiC, Shaw K, PrennushiG. 1997. The effects of humanresourcemanagementpracticeson productivity. AmericanEconomic Review 87: 291-313. Janz BD, Colquitt JA, Noe RA. 1997. Knowledge worker team effectiveness: the role of autonomy, team development,and contextual supportvariables.Personnel Psychology 50: 877-904. interdependence, VIII:User's Referenceand Guide. Scientific Software:Mooresville, IN. JoreskogKG, SorbomD. 1993. LISREL Kerr S, Hill KD, Broedling L. 1986. The first line supervisor: phasing out or here to stay? Academy of ManagementReview 11: 103-117. KomakiJL, Desselles ML, Bowman ED. 1989. Definitely not a breeze: extending an operantmodel of effective supervisionto teams. Journal of Applied Psychology 74: 522-529. Copyright? 2001 John Wiley & Sons, Ltd.

J. Organiz.Behav. 22, 537-550 (2001)

550

M. A. GRIFFIN ETAL.

Kozlowski SWJ, Doherty ML. 1989. Integrationof climate and leadership:examinationof a neglected issue. Journal of Applied Psychology 74: 546-553. LaRocco JM. 1985. Effects of job conditionson workerperceptions:ambientstimuli vs. groupinfluence.Journal of Applied Social Psychology 15: 735-757. Lawler EE, MohrmanSA, Ledford GE. 1992. Employee Involvementand Total Quality Management.JosseyBass: San Francisco. Lawthon R, Patterson M, West MA, Maitlis S. 1997. Organizational Climate Indicator. Institute of Work Psychology: Sheffield. Manz CC, Sims HP. 1987. Leading workersto lead themselves: the externalleadershipof self-managedworking Science Quarterly32: 106-128. teams. Administrative McIntyre RM, Salas E. 1995. Measuring and managing for team performance: lessons from complex environments. In Guzzo RA, Salas E. (eds). Team Effectiveness and Decision Making in Organizations. Jossey-Bass: San Francisco. Neuman G, EdwardsJ, Raju N. 1989. Organizational developmentinterventions:a meta-analysisof their effects on satisfactionand other attitudes.Personnel Psychology 48: 747-773. Ostroff C. 1992. The relationship between satisfaction, attitudes, and performance:an organizational-level analysis. Journal of Applied Psychology 77: 963-974. ParkerSK, Wall TD. 1998. Job and Work to PromoteWell-Being and Effectivness.Sage: Design: OrganizingWork London. PritchardRD. 1992. Organizational productivity.In Handbookof Industrialand OrganizationalPsychology, 2d ed, Vol.3, DunnetteMD, Hough LM. (eds). ConsultingPsychologists Press: Palo Alto, CA; 443-471. Ryan AM, Schmit MJ, Johnson R. 1996. Attitudes and effectiveness: examining relations at an organizational level. Personnel Psychology 49: 853-882. ShroutPE, Fleiss JL. 1979. Interclasscorrelations:use in assessing raterreliability.Psychological Bulletin 86: 420-428. Steers RM, Rhodes SR. 1978. Major influences on employee attendance:a process model. Journal of Applied Psychology 63: 391-407. Tannenbaum SI, Salas E, Cannon-BowersJA. 1996. Promotingteam effectiveness. In Handbookof WorkGroup Psychology, West MA. (ed.). John Wiley & Sons. Chichester,UK. 503-529. Tjosvold D. 1995. Cooperationtheory, constructive controversy,and effectiveness. In TeamEffectiveness and Decision Making in Organizations,Guzzo RA, Salas E. (eds). Jossey-Bass: San Francisco. Wall TD, Kemp NJ, Jackson PR, Clegg CW. 1986. Outcomes of autonomouswork groups: a long-term field experiment.Academy of ManagementJournal 29: 280-304. Warr PB, Cook JD, Wall TD. 1979. Scales for the measurements of some work attitudes and aspects of psychological well-being. Journal of OccupationalPsychology 52: 1063-1070. Yukl G. 1989. Leadershipin Organizations,2nd ed. Prentice-Hall:Englewood Cliffs, NJ.

Copyright ? 2001 John Wiley & Sons, Ltd.

J. Organiz.Behav. 22, 537-550 (2001)