Вы находитесь на странице: 1из 22

The International Journal of Human Resource Management, Vol. 20, No. 2, February 2009, 399–419

Management, Vol. 20, No. 2, February 2009, 399–419 Development and performance of self-managing work teams: a

Development and performance of self-managing work teams: a theoretical and empirical examination

Ben. S. Kuipers a * and Janka I. Stoker b

a Faculty of Social Sciences, Erasmus University Rotterdam, The Netherlands; b Faculty of Economics and Business, University of Groningen, The Netherlands

Several theories have been developed that prescribe the team development of self- managing work teams (SMWTs). Some of these have led to models with successive linear developmental phases. However, both the theory and the empirical data show little support for these models. Based on an extensive review of team development literature, we propose, instead of linear phases, describing team development in three general team processes. These processes, internal relations, task management, and external relations and improvement, were empirically explored in a longitudinal field- study of more than 150 blue-collar and white-collar SMWTs in a Volvo plant in Sweden. The three processes were found to be consistent over time and appeared to relate to one-year-later objective SMWT performance measures for product quality, the incidence of sick-leave and long-term sick-leave. Based on these findings, a result- oriented team development approach is proposed, in which the achieved results determine the processes followed to develop SMWTs further. Also, managers and HR practitioners are encouraged to monitor the three ongoing team processes and to relate these to the desired team performance. Such an analysis should be the starting point of a dialogue between manager and team to improve the functioning and performance of SMWTs.

Keywords: business performance; quality of working life; self-managing work teams; team development; team processes

Introduction

The use of teams has grown increasingly popular in organizations over recent decades (witnessed by the special issue of The International Journal of Human Resource Management in February 2005). Many publications in professional journals and the applied press have appeared (see O’Connell, Doverspike and Cober 2002). In their summary and review of research on teams, Cohen and Bailey (1997) define four different team types: work, parallel, project, and management. Self-Managing Work Teams (SMWTs) are a particular form of work team (Spreitzer, Cohen and Ledford 1999) and are the focus of this paper. They can be defined as groups of interdependent individuals that are able to self-regulate their behaviour concerning relatively complete tasks (Spreitzer et al. 1999). Self-management refers to a ‘reduced need for hierarchical command and control leadership’ (Morgeson 2005) in organizations. SMWTs are adopted in many organizations in order to improve performance and the wellbeing of employees (Hackman 1990; Manz and Sims 1993; Cascio 1995; Cohen, Ledford and Spreitzer 1996; Spreitzer et al. 1999).

400

B.S. Kuipers and J.I. Stoker

An important, and not necessarily straightforward, issue in achieving self-management is the development path towards this goal. The main line of thinking in several publications on this subject is that the development of SMWTs can be described in distinct linear phases (Katzenbach and Smith 1993; Zenger, Musselwhite, Hurson and Perrin 1994; Van Amelsvoort and Benders 1996). However, O’Connell et al. (2002) observe that such publications seldom contain empirical support for this statement. Although some research has vigorously investigated these prescriptive linear phases (see, for example, Miller 2003), the debate on team development in academic journals has focused more on theoretical descriptive frameworks and the taxonomies of team processes (Marks, Mathieu and Zaccaro 2001). Unfortunately, these articles are again rarely based on real, in-context, empirical data and, if they are, as is the case for the work of Gladstein (1984), they tend to be focused on common work teams rather than SMWTs. Overall, there seems to be little consensus on the overall development processes associated with SMWTs. Although several authors agree that SMWTs somehow develop towards greater self-management, and thereby achieve increased performance and enhanced quality of working life, there is no consensus on how this occurs and the type of performance-related outputs that can be expected. Moreover, there is a lack of empirical data to support or refute the various claims. Given the growing importance of SMWTs, we clearly need a better theoretical and empirical understanding of team development and its relationship to performance within SMWTs. Therefore, in this paper, we will first review a variety of phase models as well as other approaches to team development and team performance. Then, based on this, we propose an alternative view of team development processes. Following this, these team processes will be empirically related to team performance indicators (product quality and absenteeism) in a longitudinal field-study involving more than 150 SMWTs at a Swedish Volvo plant. Finally, we discuss the implications of our findings for both theory and practice.

Team development literature

Various schools of thought can be distinguished that deal with the issue of team development. In the following subsections, we will briefly discuss key literature for each of them, including their strengths and weaknesses. Overall, it seems reasonable to divide the approaches into three main types:

1. phase models (including group dynamics, consultancy and sociotechnical phase models);

2. recurring phase models; and

3. process models.

Phase models: Group dynamics

The most commonly used and cited approach in the group-dynamics literature (Miller 2003) is the group development theory by Tuckman (1965), later extended by Tuckman and Jensen (1977). This theory describes five stages through which a group passes:

1. Forming. The initial group phase of orientation among group members in which interpersonal and task behaviour is tested.

The International Journal of Human Resource Management

401

3. Norming . Overcoming the resistances of the second phase to achieve group cohesiveness and developing norms and roles occur in this third phase.

4. Performing. This is the fourth stage of group development and focuses on task performance. The roles and group structure developed in the norming phase form the basis for accomplishing the task.

5. Adjourning . In this final phase the group separates and, in a new form, starts again with the forming phase.

Tuckman built his theory using the concept of ‘interpersonal stages of group development

that any group, regardless of setting, must address

itself to the successful completion of a task. At the same time, and often through the same behaviours, group members will be relating to one another interpersonally’ (1965, p. 385). He based his successive stages of group development on an extensive literature search involving findings related to therapy groups, training groups and laboratory groups. Despite the popularity of Tuckman’s model, three fundamental criticisms have been levelled at it. The first is that team development often deviates from the sequential steps suggested (Forsyth 1999). Groups omit certain of the phases defined by Tuckman, move through the phases in a different order or develop in ways that cannot be described by these

phases (Seeger 1983). The second criticism is that it is impossible to demarcate clearly between the phases since certain group dynamical aspects do not occur timely nor in sequential order (Arrow 1997). That is, in practice, teams do not always develop according to clear, distinguishable phases. Third, the theory was based on the temporal patterns found in time-limited therapy and laboratory groups, and it has been questioned whether such patterns can adequately describe work-team processes in an organizational setting. As Cohen and Bailey (1997, p. 240) observe, ‘The findings from studies of undergraduate psychology or business students are much less likely to apply to practicing managers, employees or executives.’ The authors also noted that many of the studies involving laboratories failed to examine organizational features external to the teams.

and task behaviours’ on the ‘contention

Phase models: Consultancy practice

In terms of consultancy practice, several phase models have been developed. Katzenbach and Smith (1993), as an example, define five phases of team development in their ‘team performance curve’: the team starts out as a working group, and ends up by being a high-performance team. Another well-known best-practice model by Wellins, Byham and Wilson (1991) describes a similar method for empowering teams – by increasing levels of job responsibility and authority. Important elements of these phase models are: the development of joint accountability, the goal direction and the performance focus of the team; and these are related to the team’s group dynamical phases. Although these models are highly prescriptive and poorly defined for academic application, their role in team development practice should not be underestimated. Offerman and Spiros (2001) note that Katzenbach and Smith’s 1993 book The Wisdom of Teams is the ‘most commonly cited’ book by both ‘full-time practitioners’ and ‘academic practitioners’. One should also note that these popular teamwork phase models clearly stem from Tuckman’s model described earlier.

Phase models: Sociotechnical approach

Based on sociotechnical principles (Morgan 1993), Van Amelsvoort and Benders (1996) also developed a phase model for SMWTs. This model was inspired by the work of Katzenbach and Smith (1993) and Tuckman and Jensen (1977) but, in every phase, aspects

402

B.S. Kuipers and J.I. Stoker

of the sociotechnical concept are included. The phases (Van Amelsvoort and Benders 1996; Hut and Molleman 1998; Van Amelsvoort and Van Amelsvoort 2000; Kuipers and De Witte 2005) can be described as:

.

Phase 1 . Involves a ‘bunch of individuals’ with a focus on technical proficiency, and this leads to a broadening of the types of tasks performed. The job content is increased by focusing on the redundancy of functions and on multi-functionality. All members of the team must be able to perform the primary tasks of the team.

.

Phase 2 . The ‘group’ acquires a focus on managerial autonomy. This implies that team members are empowered through adding greater decision-making authority to their tasks, and thereby increasing the team’s responsibility. The key characteristic of this phase is also called ‘minimal critical specification’ (Morgan 1993). Managers, from production as well as from supporting departments, delegate some of their responsibilities to the team, such as for quality and planning activities.

.

Phase 3 . The ‘team’ develops a focus on social maturity, which is also described as the ‘self-reliance of the team’. The team has to work as a team, and this involves teambuilding, working on communication, and joint decision-making. The team grows in autonomy and becomes increasingly independent of its supervisor.

.

Phase 4 . The ‘open team’ with a performance focus evolves. The principles of this phase are ‘double-loop learning’ and developing the capacity to solve most non- routine problems. It concerns the ‘management of team boundaries’. This idea is based on Katz and Kahn (1978) and relates to building relationships with other teams, customers and suppliers.

Empirical support for phase models

Empirical support for the phase models discussed above is limited. For example, although Van Amelsvoort and Benders (1996) investigated 267 teams by a ‘quick-scan’, the items used in this scan and the measurement methods are not clearly explained. Further, they report that 26% of the teams were newly established, 63% were in Phase 2, 8% had entered Phase 3 but none had reached the fourth phase (De Leede and Stoker 1996). Hut and Molleman (1998) investigated the sociotechnical phase approach by integrating it with the theories of Wellins et al. (1991) and Campion, Medsker and Higgs (1993). Their article presents the outcomes of a small survey among four teams and involved measuring four successive phases. Although their sample was rather small, the results are nevertheless interesting. They show that teams cannot be positioned in a single phase at a particular time; rather, teams are developing in all four phases at the same time. Nevertheless, for three of the teams, they concluded that the first phase had been developed the most, followed by the second, the third and finally the fourth phase. This pattern of overlapping phases does suggest that teams do indeed move from simple to complex tasks. Based on the study of Hut and Molleman (1998), Kuipers and De Witte (2005) conducted a study of 37 assembly teams at a Swedish Volvo Truck plant. They also failed to recognize a pattern of phased development and, further, they could not even detect the overlapping pattern suggested by Hut and Molleman. The teams did not show any particular pattern in their development. De Leede and Stoker (1999) examined SMWTs in 11 companies, and they also failed to find the linear developments described by Katzenbach and Smith (1993) and by Van Amelsvoort and Benders (1996). They suggested that the normative character of the phase theories might partly explain this

The International Journal of Human Resource Management

403

discrepancy. In an earlier publication, De Leede (1997) argued that such models connect structural change to a group dynamical change and wondered whether the structural change transitions really took place at the exact same time as group dynamical change transitions. Overall, it is reasonable to conclude that phase models lack a firm empirical basis resulting from the study of real work teams. A final critical remark concerns the fact that, as with the consultancy practice approach, these phase models can also be traced back to the ideas of Tuckman (1965). However, the prescriptive nature of linearly developing SMWTs cannot be justified using the predominantly descriptive nature of Tuckman’s original study. Moreover, Tuckman developed his model based largely on studies concerning group development in therapy groups and laboratory settings. Ignoring the quite different nature of group working in organizations, the proponents of the contemporary phase models adopted these phases for the development of SMWTs without regard to the specific nature and characteristics of such teams. Given the differences, the use of Tuckman’s descriptive model for team development is questionable in developing a prescriptive approach for a work-related setting.

Recurring phase models

The criticisms regarding Tuckman-like successive phase theories led to another perspective on teamwork phases being developed. Gersick (1988; 1989) can be seen as one of founders of this approach which sees the developmental process of a team as much more complex than a number of sequential phases. She studied the development of groups and subsequently introduced the idea of two main phases. Her punctuated equilibrium model describes an initial phase which, half way through the group’s lifespan, undergoes a transition into a certain action phase . Gersick’s ideas have served as input to other models, such as the one by Marks et al. (2001). Both the theories of Gersick (1988; 1989) and of Marks et al. (2001) can be labelled as recurring phase models , with transaction and action phases taking turns through time for the various tasks and sub-tasks. Marks et al. (2001, p. 357) define team processes as ‘members’ interdependent acts that convert inputs to outcomes through cognitive, verbal, and behavioural activities directed toward organizing task work to

’. Their descriptive approach is based on the ‘idea that teams

perform in temporal cycles of goal-directed activity, called “episodes”

achieve collective goals

’ (p. 359). They

also place an emphasis on interpersonal processes occurring ‘throughout both transition and action phases, and typically lay the foundation for the effectiveness of other

’ (p. 368). They describe ten sub-processes that can take place, and these are

processes

allocated to the two episodes and to the associated interpersonal processes. The transition phase consists of the mission analysis, goal specification, and strategy formulation and planning sub-processes. The action phase includes monitoring progress toward goals, systems monitoring, team monitoring and backup, and coordination sub-processes. The interpersonal processes are ‘conflict management’, ‘motivation and confidence building’, and the ‘affect management’ sub-processes (Marks et al. 2001).

Process models

Another, more process-oriented, theory linking teamwork to performance is Gladstein’s concept of group processes (1984). In her study of 100 small sales teams (2 – 4 people), she showed that the group processes she was measuring were clearly dividable between an

B.S. Kuipers and J.I. Stoker404

Table 1. Summary of linear phase models.

Group dynamic phase models Consultancy phase models

Sociotechnical phase models

 

Tuckman and Jensen

Wellins, Byham and Wilson (1991)

Katzenbach and Smith

Van Amelsvoort and Benders (1996)

Hut and

Comparable

Phase

(1977)

(1993)

Molleman (1998)

processes

Phase 1

Forming: Orientation

Diverse collection of individuals, arranging simple tasks

Combination of individuals changing information and successful working methods No performance focus

Multi-skilling

Multi-functionality

Task

and testing

 

management

(task) behaviour

 

Phase 2

Focus on tasks and roles, increasing responsibilities

Common tasks and goals, influence on performance improvement

Managerial tasks

Control tasks

 

Storming: Conflicts and group influence

 

Internal

relations

Phase 3

Norming: Cohesiveness, developing norms and roles

Common processes and crisis situations

Collective accountability

Team-building and

Self-reliance

collective goals

and team

 

communication

Phase 4

Performing: Accomplishing the task

Continuous improvement, proactiveness, arranging complex tasks and higher order responsibilities

Deep commitment to mutual growth and success

External team relationships and high performance

Boundary management

External

and

relations and

   

non-routine problems

improvement

Phase 5

Adjourning: Separation

The International Journal of Human Resource Management

405

Table 2. Literature overview of aspects of team development and the relation with team results.

Key aspects

Characterization

Authors

Model type

Performance orientation

Internal relations

Goal orientation

Determining team goals

Katzenbach and Smith (1993), Marks et al. (2001), Wellins et al. (1991)

2,4

Self-reported effectiveness, work satisfaction, customer satisfaction (Gladstein 1984); value creation (Dunphy and Bryant 1996)

Planning activities

Team planning of work and support activities

Wellins et al. (1991), Dunphy and Bryant (1996), Van Amelsvoort and Benders (1996), Hut and Molleman (1998) Marks et al. (2001), Gladstein (Gladstein 1984), Hut and Molleman (1998)

2,3,4

Feedback

Motivation, assessment and constructive feedback in task performance Handling cooperation and behaviour problems

3,4,5

Conflict management

Marks et al. (2001), Gladstein (1984)

4,5

Task management

 

Multi-functionality

Task flexibility and appli- ance of multi-skilling

Dunphy and Bryant (1996), Van Amelsvoort and Benders (1996), Hut and Molleman (1998)

3,5

Affective and behavioural responses (Dunphy and Bryant); work satisfaction (Gladstein 1984), individual performance, costs, value creation (Dunphy and Bryant

 

1996)

Delegated management and support tasks

Carrying out and arranging routine production support activities Sharing strictly task-related information Joint performance of man- agerial tasks

Wellins et al. (1991), Dunphy and Bryant (1996), Van Amelsvoort and Benders (1996), Hut and Molleman (1998) Gladstein (1984), Wellins et al. (1991), Van Amelsvoort and Benders (1996) Wellins et al. (1991), Dunphy and Bryant (1996), Van Amelsvoort and Benders (1996), Hut and Molleman (1998)

2,3,5

Work communication

2,3,5

Decision-making and control

2,3,5

B.S. Kuipers and J.I. Stoker406

Table 2 – Continued

Key aspects

Characterization

Authors

Model type

Performance orientation

Performance management

Actions to improve the team’s performance

Katzenbach and Smith (1993), Marks et al. (2001), Wellins et al. (1991), Dunphy and Bryant (1996), Van Amelsvoort and Benders (1996), Hut and Molleman (1998)

2,3,4,5

External relations and improvement Improvement activities

Initiating and supporting product and process improvements

Katzenbach and Smith (1993), Marks et al. (2001), Wellins et al. (1991), Dunphy and Bryant (1996), Van Amelsvoort and Benders (1996), Hut and Molleman (1998)

2,3,4,5

Self-reported effectiveness, work satisfac- tion, customer satisfaction (Gladstein 1984); innovation (Dunphy and Bryant 1996); ‘high performance’ (Van Amels- voort and Benders 1996; Wellins et al.1991; Katzenbach and Smith 1993)

Customer and supplier relations

Maintaining relations with internal and external customers

Gladstein (1984), Katzenbach and Smith (1993), Wellins et al. (1991), Van Amelsvoort and Benders (1996), Hut and Molleman (1998) Wellins et al. (1991), Dunphy and Bryant (1996), Van Amelsvoort and Benders (1996), Hut and Molleman (1998)

2,3,5

Advanced management and support activities

Carrying out and arranging non-routine production support activities

2,3,5

Notes: 2 Consultancy phase models; 3 Sociotechnical phase models; 4 Recurring phase models; 5 Process models

The International Journal of Human Resource Management

407

intra-group process and a boundary management process . The first included aspects such as open communication, supportiveness, conflict management and discussion of strategies. The concept of boundary management on the other hand is defined as the ‘degree of misunderstanding with external groups’ (Gladstein 1984). She stressed the importance of the difference between the two types of processes: ‘Clearly, in organisational settings many groups cannot work in the isolation enjoyed by groups in a laboratory setting. These ’

groups need to manage their boundaries and adapt to their organizational environment

(Gladstein 1984, p. 513). A feature of Gladstein’s theory is that it sets out to describe the processes occurring within teams without trying to order what comes first and what comes last. She also considers the intra-group processes and boundary management to be parallel processes. Similarly, Dunphy and Bryant (1996) define three team attributes that they see as ‘creating an agenda for team development’: 1) technical expertise ; 2) self-management ; and 3) self-leadership. Although these attributes are not directly defined as processes, they can easily be regarded as such. Under technical expertise , the members of a team work on multi-skilling for an expanded task. Self-management concerns the delegation of ‘operational responsibilities’ from the manager to the team. Self-leadership involves

elements of both cooperation and continuous improvement. Teams that have developed this final attribute are seen as the self-governing basic units of the organization, they can play a strategic role and can provide improved communications both within and beyond the team (Dunphy and Bryant 1996). The authors link their team attributes to three sets of performance outcomes: costs; value; and innovation. In a third and related approach, Kuipers and De Witte (2005) suggest considering the simultaneous processes that occur during the life of an SMWT. Based on their empirical

finding that there were ‘very few teams [that] exhibited a linear pattern

the existence of parallel dimensions that could be developed independently of each other.

’, they proposed

Patterns in team development theories

From the above discussion, based on a literature review, it can be concluded that several theories adhere to some form of linear phase approach. Further, as can be seen in Table 1, there are clear similarities between the various linear phase models. The overview provided in Table 1 shows that the various linear phase models can indeed all be seen as refinements of the original model by Tuckman (1965). Further, all the models indicate, in one way or another, stepwise growth in performance with each successive development phase, resulting in some final ‘high-performance’ phase. If we take this analysis one step further, and compare not only the linear phase models but also the recurring phase models and the process models, it seems that there are three team processes that essentially cover all the theoretical ideas expounded. In Table 2, these three processes are described and linked to the related models by determining 12 key aspects. First, if we look at how these three processes resemble the phase models, the process of task management generally covers the first two phases of the models, i.e. activities linked to team multifunctionality and its capabilities to manage responsibilities and control. Next, the process of internal relations covers the third phase of the consultancy and sociotechnical phase models and the second phase of Tuckman’s model, in which the team deals with internal cooperative issues. Finally, the external relations and improvement process covers the fourth phase of all the models. In this process, the team deals with its relationships with other teams, customers and suppliers and works on improving performance.

408

B.S. Kuipers and J.I. Stoker

Further, these three processes can also be clearly related to the recurring phase models and the process models. After all, Marks et al. (2001) refer to the team’s interpersonal processes and Gladstein (1984) distinguishes the intra-group process from a team’s external relationships. Task management can be found in Dunphy and Bryant’s technical expertise and self-management attributes (1996), and Marks et al. (2001) are referring to this topic in their action phase. In terms of the third process, Dunphy and Bryant (1996) see the team’s external relations as part of self-leadership, and Gladstein (1984) uses the specific terminology of boundary management, which she defined as the degree of misunderstanding between the team and external individuals and groups. Based on this assessment, we argue that these three processes are sufficient to describe team development. The process of internal relations refers to the internal cooperation and common accountability of the team. Internal relations include all the activities that potentially connect the members as a team, such as goal orientation and planning activities, as well as the relational processes of feedback and conflict management. Second, the process of task management represents the extent to which the team manages its primary process. It includes aspects of both job enlargement and job enrichment, such as multifunctionality, delegated management support tasks, decision-making and control. It also encompasses basic work communication and performance management. The final process of external relations and improvement reflects the extent to which the team explores and develops its boundaries and, as such, it is broader than Gladstein’s (1984) concept of boundary management. Here, in addition to customer and supplier relations, we include improvement activities and the team’s advanced managerial and support function.

Team development and performance

We argue that the above three processes offer a suitable and all-embracing perspective on team development. However, one issue is still missing, namely the explicit relationship with performance. Team development should not be regarded as a goal in itself (see Kuipers 2005), since teamwork is intended to achieve organizational goals: i.e. a team is a means to an end rather than an end in itself. That is, the development of SMWTs should be aimed at improving results, both in terms of organizational performance and in terms of the wellbeing of employees. The contribution here by authors such as Gladstein (1984), Dunphy and Bryant (1996) and Kuipers and De Witte (2005) is that they make a connection between the development (processes) of teams and the team performance. Thus, the elaboration of the above three team processes would be incomplete without relating them to specific SMWT performance outcomes. Since team development is not a goal in itself, but a means to achieve certain desired outcomes, we will define team development in this paper, following Kuipers (2005), as the overall set of group processes reflecting a team’s actions and behaviour to given tasks, goals and challenges, resulting in desired outcomes of teamwork . Each of the theories introduced earlier in this paper describes the performance of a team, implicitly or explicitly, as a result of one or more of the stages or attributes that a team passes through. While Katzenbach and Smith’s learning performance curve (1993) and Dunphy and Bryant’s (1996) connection between team attributes and team performance are the clearest examples of this general situation, even Tuckman (1965) calls one of the team stages ‘performing’. However, in general, the authors remain vague as to the performance level that might be achieved. Further, there is very little empirical evidence supporting the relationship between team development and performance.

The International Journal of Human Resource Management

409

Table 3. Measures for team processes.

Key aspects

No. of items

Example item

Cronbach’s alphas*

Internal relations

Goal orientation

2

Team goals are formulated by the team and based on the company’s goals The team formulates its own weekly production plan The team members address to each other in case of mistakes in the task performance The team members solve internal cooperation problems without management interference

.76, .76, .74

Planning activities

4

.79, .76, .78

Feedback

3

.70, .63, .67

Conflict management

4

.78, .77, .81

Task management Multifunctionality Delegated management and support tasks

5

4

The team members often interchange tasks The team carries out the routine maintenance The team members share information about the work The team divides the tasks The team acts on mistakes

.78, .77, .79 .74, .75, .75

Work communication

2

.64, .70, .66

Decision making and control Performance management

6

.76, .84, .86 .68, .67, .70

3

External relations and improvement Improvement activities

4

The team members often take initiatives for improvement The team solves problems with internal customers The team arranges back-up and support when necessary

.76, .76, .77

Customer and supplier relations

4

.87, .86, .88

Advanced management and support activities

5

.69, .72, .73

Note: * The sample size ranges between n ¼ 1,293 (2001) and n ¼ 1,507 (2002).

410

B.S. Kuipers and J.I. Stoker

In general, one can differentiate between two types of team performance (see Dunphy and Bryant 1996; Kuipers and De Witte 2005). One reflects the quality of working life (QWL), referring to the wellbeing of people in organizations in terms of satisfaction (a subjective measure), involvement and also absenteeism and sick leave (objective measures). The other is business performance (BP), also known as organizational performance, which is addressed using objective indicators such as product quality, productivity, costs and delivery precision. Researchers have complained about the lack of available studies on the latter (Dunphy and Bryant 1996) and it is claimed that it is difficult to obtain appropriate objective outcome measures (Parker 2003). Nevertheless, empirical studies using appropriate measures are important to reveal the potentials of teamwork in terms of performance (Dunphy and Bryant 1996). Moreover, if one is not only to relate team development and outcomes, but also to explain the causalities, longitudinal data are essential. One of the general limitations of earlier empirical studies is that they often lacked longitudinal data (Kozlowski and Bell 2003) and, as a consequence, their cross-sectional design precludes any conclusions about the direction of causality. From these findings, we concluded that insights into the longitudinal relationship between team development processes and the objective measures for both QWL and BP were needed. In this paper, therefore, we will describe results of a longitudinal study looking into both aspects. We set out to explore the relationships between the three processes of internal relations, task management, and external relations and improvement on the one hand, and team performance measures on the other. Based on the limited available literature, we hypothesized the following relationships. First, regarding internal relations , Marks et al. (2001, p. 368) state that especially

typically lay the foundation for the effectiveness of other

processes’. The relationships found by Gladstein (1984) between intra-group processes and self-reported group effectiveness provide additional support to this idea. Campion et al. (1993, p. 841) similarly report ‘the importance of proper group processes to the functioning of effective work groups’. In line with these arguments, we therefore expect to find positive relationships between the internal relations of SMWTs and primarily the QWL aspects of teams. That is, the relational aspects and interpersonal processes in teams may primarily affect the team atmosphere and the wellbeing of team members. Maintaining good internal cooperation may prevent absenteeism and support team members in returning to work sooner after sick-leave. Second, in terms of task management , job-related aspects have historically always been connected with QWL. Hackman and Oldham (1980) reported positive consequences of job enlargement and job enrichment in terms of motivation, quality of work, satisfaction and turnover, and this has been supported by other studies. For example, Yeatts and Hyten report that especially enriched work environments consistently show positive effects on employee satisfaction (1998, p. 249). Similarly, Parker (2003) showed, in a longitudinal study, how lower ‘job autonomy, skill utilisation and participation in decision making’ have negative effects on the wellbeing of employees. In other words, developments in task management are primarily expected to affect QWL (see also Table 2). Finally, we would expect external relations and improvement aspects to affect both BP and QWL since, in the literature, these aspects are often related to high performance (Wellins et al. 1991; Katzenbach and Smith 1993; Van Amelsvoort and Benders 1996), customer satisfaction (Gladstein 1984) and innovation (Dunphy and Bryant 1996). In general, the idea of high-performance reflects all the various types of performance involved. However, more specifically, it is the direct customer and supplier relationships

‘interpersonal processes

The International Journal of Human Resource Management

411

Table 4. Data on business performance and quality of working life (average per team).

 

2002

2002 2003

Product quality (Blue-collar teams)

n

73

M

91.96

s.d.

4.79

Short term sick-leave (Blue-collar and white-collar teams)

n

57

M

2.39

s.d.

1.70

Long-term absenteeism (Blue-collar and white-collar teams)

n

57

M

1.25

s.d.

2.43

Notes: n ¼ number of teams; M ¼ mean; s.d. ¼ standard deviation.

that provide inputs to improve products and processes. Working on these relationships may lead to better quality products being produced and also to a healthier working environment. People may feel more committed and involved through managing good relationships with customers and suppliers and therefore receive more intrinsic motivation to be present at work and ‘help the customer’. As a consequence, absenteeism may be reduced.

Method and measures

At one-year intervals, three sets of data were collected using questionnaires distributed among 2,200 employees working in more than 150 SMWTs at a Volvo Trucks plant in Sweden by the first author (see Kuipers 2005). Response rates for the 2001, 2002 and 2003 exercises were 73%, 76% and 68%, with data obtained from 152, 168 and 167 teams respectively. The vast majority of the teams were production teams (about 75% of the total) working in one of the plant’s five production departments (identified as blue-collar teams). The other teams were related to service and support, ranging in function from engineering to financial administration (white-collar teams). Conventional paper-based questionnaires were used, and these were distributed by the team managers and completed during working hours at one of the weekly team meetings. Volvo itself provided the objective measures that were used for both BP and QWL. Unfortunately, the BP data were available only for blue-collar teams, whereas QWL data were available for both blue- collar and white-collar teams. Initially, we sought confirmation of the existence of the three team processes proposed. To achieve this, the data for all the teams from the three questionnaire rounds were compared using a factor congruency test. Next, the relationships between the three processes and BP and QWL were established using longitudinal regression models: the scores for the team processes in 2001 serve as input to the BP measures in 2002 and to the average QWL measures over the period 2002– 2003. The sample size for these latter analyses is dependent on the availability of data (see below). The questionnaire consists of 46 self-reported Likert-type items (1 ¼ ‘strongly disagree’, 5 ¼ ‘strongly agree’) based on the aspects mentioned in Table 1 and originating from Hut and Molleman (1998), the Work Groups Effectiveness Model by Campion et al. (1993), De Leede and Looise (1999) and Kuipers and De Witte (2005). These items (see Table 3) cover the 12 previously defined key aspects. Table 3 provides an overview of the number of items used for each aspect, plus example items and Cronbach’s Alpha calculated for each aspect (which indicates adequate reliability) for each year of

412

B.S. Kuipers and J.I. Stoker

Table 5. Regression results for one-year-later BP and QWL measures.

 

BP: Product

QWL: Number of sick-occasions b -values

QWL: Percentage of long-term absenteeism

quality

Control: Blue-collar teams n.a.

 

.603 *** 2 .057 2 .145 2 .394 ** 42 (4)

.311

Internal relations

2 .019

2 .368 * 2 .086

Task management

.311

**

External relations

.308 **

.017

df

47 (3)

42 (4)

R 2

.279

.794

.211

Notes: * p , .05; ** p , .01; *** p , .001.

measurement. The items specifically reflect actions and behaviour of teams, that is, none

of the items refer to emergent states (Marks et al. 2001) or to actions and behaviours of

individual team members. Three objective performance measures were also used, and these are presented in Table 4. One of these represents business performance (product quality) and the other two quality of working life (number of sick-occasions and long-term absenteeism). The data for these were provided by Volvo. Unfortunately, these performance measures were not available for all teams. The measure used for product quality was only applicable to production teams. The measure is the percentage of so-called Direct OK items produced – that is the percentage of products that are produced fault-free at the first attempt by a particular team. This figure is an average measure of the product quality delivered by a team, and the data used here reflects the week in which the questionnaire was answered. In terms of sick leave, Volvo provided data on the number of sick occasions and the percentage of long-term sick leave in the organization. The data for sick leave were only available as annual averages but at least these periods started on the same date as the questionnaire was answered. Both these measures were available for both blue-collar and white-collar teams. Product quality data were available for 73 teams (43% of the total number of teams) and data

about sick leave were available for 57 teams (34% of the total number of teams).

Results

Factor analysis of team development

A factor analysis (principal components) was carried out, for each year of measurement,

on the 46 items related to team processes. The varimax rotated solution included three

factors (see Appendix). A factor congruency test was carried out using the formula of Gorsuch (1974) in order to see if the three factors were similar for each year. The outcomes

of 2001 were compared with those of 2002, those of 2001 with 2003 and the outcomes of

2002 with those of 2003. The outcomes for these three tests had congruencies of .98, .96

and .98 respectively; clearly above the .9 threshold for claiming factor replication (Gorsuch 1974). The three factors each explain between 10% and 20% of the variance for each year, with a combined total explained variance of between 41% and 43% in each year studied. Further, a Cronbach Alpha analysis of reliability showed that each of the three

item-scales had a satisfactory score above 0.8, for each of the years (see Appendix). The three factors found by the factor analysis confirm the previously defined team processes.

It is notable that an inspection of the outcomes for the teams did not suggest any linear

pattern in the development of the three processes, a similar finding to those of Kuipers and

The International Journal of Human Resource Management

413

De Witte (2005). Rather, each team seemed to have its own strengths and weaknesses, further supporting the idea of parallel processes.

Effects on team performance

The results of the regression analysis shown in Table 5 indicate the relationships between team processes and team performance. 1 First, as expected, internal relations do have a significant relationship with QWL, in the form of long-term absenteeism. Internal relations seems to be the only team process that has a significant effect here: high scores for internal relations are correlated to low scores for long-term absenteeism one-year later. Second, the level of task management in 2001 positively relates to BP (product quality) in the following year. Apparently, task management has a strong effect on this measure ( b ¼ .311 ** ). Third, as we expected,

external relations and improvement is also positively related to BP. Further, it also seems

to have a significant negative relationship with QWL, in particular the number of absences

due to sickness. We found that higher scores for external relations are related to lower one-year-later sickness frequencies. No other significant relationships were found for this sick-leave measure.

Conclusions and discussion

Conclusions

In this paper, some key theories concerning team development have been described and the criticisms levelled at them discussed. By identifying related issues that were found across all the common theories, a model was suggested that could characterize team development through three processes involving behaviour and team action. Further, we also cautioned that the use of phase approaches (e.g. Tuckman 1965; Wellins et al. 1991; Katzenbach and Smith 1993; Van Amelsvoort and Benders 1996) might easily result in team development being considered as a goal in itself, with passing through the stipulated steps becoming the goal. We argued that team development should be considered as a means to achieve better team results, both in terms of business performance (BP) and quality of working life (QWL). Our results first confirmed the existence of three team processes, internal relations , task management and external relations and improvement . Their existence was supported by data from more than 150 production and service SMWTs at a Volvo plant in Sweden. They were then tested to see if they had a longitudinal relationship with the objective team performance indicators of BP and QWL. The results showed how each of the processes has its own distinct effects on team performance. These empirical results emphasize the importance of team development for long-term team performance.

We found that internal relations relates negatively to long-term absenteeism.

A

possible explanation for this is that internal relations and the group’s attitude might help

to

prevent long-term sick leave because team members pay attention to each other. For the

same reason, it might also help in avoiding short-term sick leave turning into long-term absenteeism, if there is a feeling that the team cares for you and you care for the team. Second, task management positively relates to product quality. One explanation could be that SMWT craftsmanship depends on job management and so good task management results in the team delivering higher quality products. We did not find a significant relationship between task management and our measures for QWL. The reason that earlier studies did may be related to the fact they focused more on the effects of autonomy, job

414

B.S. Kuipers and J.I. Stoker

enlargement and job enrichment as an emergent state, or given organizational characteristic, rather than, as we did, on an aspect of team development. We chose to define task management explicitly in terms of concrete behaviour and actions, rather than as given working characteristics, and this might have led to the different findings. Further, the use of only sick-leave measures as indicators of QWL might be too limited; measures such as staff turnover perhaps need to be taken into account to fully describe QWL. Third, external relations and improvement was found to be positively related to product quality and negatively to the frequency of sick leave. Perhaps maintaining customer – supplier relationships makes people more committed to work and more reluctant to stay at home if they feel unwell. Maybe, they feel responsible, rather than pressured, to be at work to satisfy customer needs. Further, process improvements, initiated by the team through external relations, may also contribute to a higher product quality and to a healthier working place. Theoretically, this study supports the idea that the various team processes occur simultaneously as teams develop. In line with the thinking of Dunphy and Bryant (1996), this also implies that, for different aspects of performance, different accents in team development are required. Whereas phase theories specifically state that team development should go through the full range of phases, our results show that high performance levels can be reached by different combinations and patterns of the team processes. Our results also suggest a revision of the existing theories rather than yet another new theory on team development. The existing theories are able to deliver important input for a model that provides a more dynamic view on developmental processes in teams. Unfortunately, many of the models we reviewed lack a thorough empirical basis in real organization settings. Some academics have also complained about a lack of studies that include objective measures (e.g. Dunphy and Bryant 1996), or about the difficulty in gaining access to such data of good quality (Parker 2003). In this study, we avoided these problems by collecting data from a large number of teams over a three-year period and by including objective performance measures. The insights from this study should support researchers in further longitudinal studies of processes in SMWTs in organizations, and we would urge them to use objective performance measures, and not see team development as a goal in itself.

Limitations of this study

This study, apart from the theoretical examination of team development theories, had an explorative empirical nature. We found strong empirical support for the existence of three team processes, although further study is necessary since we faced difficulties in trying to hypothesize the longitudinal relationships between these processes and objective team performance measures. Very little literature is available on these relationships, especially in terms of having both objective and longitudinal data, and this restricts ones ability to hypothesize with confidence. To make a more general application of the model possible, further research is needed that includes additional objective QWL and BP performance indicators. Organizations are also interested in reducing costs, improving productivity and delivery precision and it is important for them to know if, and how, internal relations , task management and external relations effect such other team performances. Further research should also involve other types of organizations. Despite the fact that different types of teams were involved in this research, from various types of production departments and from supporting departments, a more general approach to team development will require studies in other settings such as healthcare and commercial service industries.

The International Journal of Human Resource Management

415

Implications for management and HR practice

Contemporary approaches such as the HR scorecard (Paauwe 2004) can support practitioners in further developing our perspective on team development. The essence of approaches like the scorecard is to help managers define the connections, or causal chains, between performance and whatever is driving and enabling this performance, and to define these in terms of concrete practices (Becker, Huselid and Ulrich 2001). Having found credible evidence for the existence of a few such connections between team processes and performances, we would propose a similar approach for the development of self- management in work teams. That is, first define the performances you want to achieve and then go back and consider the team processes that enable these performances. This alternative approach should help managers to develop their teams in a more sophisticated way. HR practitioners could support managers through coaching, providing HR tools and their knowledge of organizational development. In practical terms, this means that teams should not be stimulated to develop in a structured and predefined linear order. Rather, a team’s development should be continuously evaluated by its manager, with support from a HR practitioner. Together, they can analyse the results and subsequently put an emphasis on those dimensions that match the vulnerability of the team. Teams also do not need to be fully developed in one dimension before shifting attention to another dimension. Rather, the circumstances and the specific targets and goals related to the team’s purpose should be used to define which dimensions, and which aspects within these dimensions, should be the focus in seeking to improve performance. To conclude, we stress our argument that team development is situational, and does not depend on sequential phases as phase models suggest. This insight should encourage managers and their HR practitioners to monitor ongoing team processes and relate these to the desired performance. Such a diagnosis should be the basis of a dialogue between manager and team to improve the functioning and performance of SMWTs. By studying further the SMWT development processes in real organizational settings, researchers may encourage practitioners to develop healthy scepticism of the popular prescriptive phase theories.

Note

1. Please note that the regression models for absenteeism included a statistical control for any effect of blue-collar versus white-collar SMWTs. Only in the case of the number of sick- occasions did it appear that there was a significant difference between the two types of team. Overall, it seemed that blue-collar workers were most likely to report in sick but, in terms of long-term absenteeism, no significant difference was found.

References

Arrow, H. (1997), ‘Stability, Bistability, and Instability in Small Group Influence Patterns,’ Journal of Personality and Social Psychology , 72, 1, 75 – 85. Becker, B.E., Huselid, M.A., and Ulrich, D. (2001), The HR Scorecard, Linking People, Strategy, and Performance , Boston, MA: Harvard Business School Press. Campion, A.M., Medsker, G.J., and Higgs, A.C. (1993), ‘Relations between Work Group Characteristics and Effectiveness: Implications for Designing Effective Work Groups,’ Personnel Psychology , 46, 823– 850. Cascio, W.F. (1995), ‘Whither Industrial and Organizational Psychology in a Changing World of Work?,’ American Psychologist , 50, 928– 939.

416

B.S. Kuipers and J.I. Stoker

Cohen, S.G., and Bailey, D.E. (1997), ‘What Makes Teams Work: Group Effectiveness Research from the Shop Floor to the Executive Suite,’ Journal of Management , 23, 3, 239– 290. Cohen, S.G., Ledford, G.E., and Spreitzer, G.M. (1996), ‘A Predictive Model of Self-Managing Work Team Effectiveness,’ Human Relations, 49, 643– 680. De Leede, J. (1997), Innoveren van onderop: Over de bijdrage van taakgroepen aan product- en procesvernieuwing (Bottom-Up Innovation: About the Contribution of Self-Managing Teams to Product and Process Improvement) , Enschede, The Netherlands: University of Twente. De Leede, J., and Looise, J.K. (1999), ‘Continuous Improvement and the Mini-Company Concept,’ International Journal of Operations and Production Management, 19, 11, 1188– 1202. De Leede, J., and Stoker, J.I. (1996), ‘Taakgroepen in de Nederlandse industrie: e´ e´ n concept met vele toepassingen (Work Teams in the Dutch Manufacturing Industry: One Concept with Many Applications),’ Tijdschrift voor Arbeidsvraagstukken , 12, 4, 310– 321. De Leede, J., and Stoker, J.I. (1999), ‘Self-Managing Teams in Manufacturing Companies:

Implications for the Engineering Function,’ Engineering Management Journal , 11, 3, 19 – 24. Dunphy, D., and Bryant, B. (1996), ‘Teams: Panaceas or Prescriptions for Improved Performance,’ Human Relations , 49, 5, 677– 699. Forsyth, D.R. (1999), Group Dynamics (3rd ed.), Belmont: Wadsworth/ITP. Gersick, C.J.G. (1988), ‘Time and Transition in Work Teams: Toward a New Model of Group Development,’ Academy of Management Journal, 31, 1, 9 – 41. Gersick, C.J.G. (1989), ‘Marking Time: Predictable Transitions in Task Groups,’ Academy of Management Journal , 32, 2, 274– 309. Gladstein, D.L. (1984), ‘Groups in Context: A Model of Task Group Effectiveness,’ Administrative Science Quarterly , 29, 4, 499– 517. Gorsuch, R.L. (1974), Factor analysis , Philadelphia, PA: Saunders. Hackman, J.R. (1990), Groups That Work (and Those That Don’t): Creating Conditions for Effective Teamwork (2nd ed.), San Francisco, CA: Jossey Bass. Hackman, J.R., and Oldham, G.R. (1980), Work Redesign , Boston, MA: Addison-Wesley. Hut, J.A., and Molleman, E. (1998), ‘Empowerment and Team Development,’ Team Performance Management , 4, 2, 53 – 66. Katz, D., and Kahn, R.L. (1978), The Social Psychology of Organizations (2nd ed.), New York: John Wiley and Sons. Katzenbach, J.R., and Smith, D.K. (1993), The Wisdom of Teams; Creating the High-Performance Organization, Boston, MA: Harvard Business School Press. Kozlowski, S.W.J., and Bell, B.S. (2003), ‘Work Groups and Teams in Organizations,’ in Handbook of Psychology: Industrial and Organizational Psychology (Vol. 12), eds. W.C. Borman, D.R. Ilgen and R.J. Klimoski, Hoboken, NJ: John Wiley and Sons, pp. 333– 375. Kuipers, B.S. (2005), Team Development and Team Performance. Responsibilities, Responsiveness and Results: A Longitudinal Study of Teamwork at Volvo Trucks Umea˚ , PhD, University of Groningen. Kuipers, B.S., and De Witte, M.C. (2005), ‘Teamwork: A Case Study on Development and Performance,’ International Journal of Human Resource Management , 16, 2, 185– 201. Manz, C., and Sims, H. (1993), Business Without Bosses: How Self-managing Teams Are Building High Performance Companies , New York: Wiley. Marks, M.A., Mathieu, J.E., and Zaccaro, S.J. (2001), ‘A Temporally Based Framework and Taxomony of Team Processes,’ Academy of Management Review , 26, 3, 356– 376. Miller, D.L. (2003), ‘The Stages of Group Development: A Retrospective Study of Dynamic Team Processes,’ Canadian Journal of Administrative Sciences , 20, 2, 121– 134. Morgan, G. (1993), ‘Organizational Choice and the New Technology,’ in The Socio-Technical Perspective , eds. E. Trist and H. Murray, Philadelphia: University of Pennsylvania Press, pp. 354– 368. Morgeson, F.P. (2005), ‘The External Leadership of Self-Managing Teams: Intervening in the Context of Novel and Disruptive Events,’ Journal of Applied Psychology , 90, 3, 497– 508. O’Connell, M.S., Doverspike, D., and Cober, A.B. (2002), ‘Leadership and Semi-Autonomous Work Team Performance,’ Group and Organisation Management, 27, 1, 50 –65. Offerman, L.R., and Spiros, R.K. (2001), ‘The Science and Practice of Team Development:

Improving the Link,’ Academy of Management Journal, 44, 2, 376– 392. Paauwe, J. (2004), HRM and performance; Achieving Long Term Viability, Oxford: Oxford University Press.

The International Journal of Human Resource Management

417

Parker, S.K. (2003), ‘Longitudinal Effects of Lean Production on Employee Outcomes and the Mediating Role of Work Characteristics,’ Journal of Applied Psychology , 88, 4, 620– 634. Seeger, A.J. (1983), ‘No Innate Phases in Group Problem Solving,’ Academy of Management Review , 8, 683– 689. Spreitzer, G.M., Cohen, S.G., and Ledford, G.E. (1999), ‘Developing Effective SMWTs in Service Organizations,’ Group and Organization Management , 24, 3, 340– 366. Tuckman, B.W. (1965), ‘Development Sequences in Small Groups,’ Psychological Bulletin , 63, 6, 384– 399. Tuckman, B.W., and Jensen, M.A.C. (1977), ‘Stages of Small Group Development Revisited,’ Group and Organization Studies , 2, 419– 427. Van Amelsvoort, P., and Benders, J. (1996), ‘Team Time: A Model for Developing Self-Directed Work Teams,’ International Journal of Operations and Production Management , 16, 2, 159– 170. Van Amelsvoort, P., and Van Amelsvoort, G. (2000), Designing and Developing Self-Directed Work Teams , Vlijmen: ST-Groep. Wellins, R.S., Byham,W.C., andWilson, J.M. (1991), Empowered Teams: Creating Self-Directed Work Groups that Improve Quality, Productivity and Participation, San Fransisco, CA: Jossey-Bass. Yeatts, D.E., and Hyten, C. (1998), High-Performing Self-Managed Work Teams: A Comparison of Theory , Thousand Oaks, CA: Sage. Zenger, J.H., Musselwhite, E., Hurson, K., and Perrin, C. (1994), Leading Teams: Mastering the New Role , Homewood, IL: Business One Irwin.

B.S. Kuipers and J.I. Stoker418

Appendix 1. Factor analyses of team responsiveness dimensions (rotated factor loadings in absolute values).

 

2001

2002

2003

Factor 1

Factor 2

Factor 3

Factor 1

Factor 2

Factor 3

Factor 1

Factor 2

Factor 3

Items

Task mgt.

External rel.

Internal rel.

Task mgt.

External rel.

Internal rel.

Task mgt.

External rel.

Internal rel.

task01

.598

.485

.424

.505

.356

task02

.562

.460

.386

.458

.352

task03

.340

.451

.438

.468

task04

.647

.502

.564

task05

.596

.447

.458

task06

.706

.513

.333

.563

.359

task07

.673

.635

.653

task08

.630

.631

.676

task09

.370

.636

.662

task10

.617

.566

.331

.580

.316

task11

.556

.467

.385

.512

.365

task12

.500

.323

.474

.323

.487

.345

task13

.517

.348

.504

.362

.496

.328

.385

task14

.494

.672

.700

task15

.458

.304

.573

.581

task16

.591

.633

.664

task17

.482

.539

.522

task18

.491

.540

.599

task19

.526

.383

.380

.352

.499

.406

task20

.545

.336

.462

.367

.602

task21

.478

.360

.489

.404

.580

.370

ext01

.711

.686

.701

ext02

.829

.807

.827

ext03

.709

.686

.726

ext04

.838

.807

.825

ext05

.642

.657

.659

ext06

.702

.686

.698

ext07

.379

.541

.481

ext08

.596

.671

.613

The International Journal of Human Resource Management

419

ext09

.417

.426

.393

.434

.538

.388

ext10

.464

.398

.483

.457

.357

ext11

.429

.365

.410

ext12

.389

.463

.308

.327

.462

.337

.416

.452

ext13

.390

.459

.350

.465

.330

.403

.452

.337

int01

.577

.319

.330

.435

int02

.772

.406

.318

.480

int03

.716

.452

.345

.438

int04

.411

.448

.498

int05

.568

.361

.326

.631

.376

.612

int06

.483

.373

.339

.478

.422

.403

int07

.493

.380

.665

.689

int08

.460

.375

.609

.316

.651

int09

.442

.424

.689

.701

int10

.321

.499

.371

.432

.371

.328

.350

int11

.479

.493

.343

.547

.494

.424

int12

.431

.542

.615

a

.9054

.8724

.8641

.9179

.8832

.8480

.9274

.8878

.8648

Notes: Extraction method: Principal Component Analysis; Rotation method: Varimax with Kaiser normalization; Factor loadings smaller than .30 are suppressed; Loadings in italic indicate to which factor the item was assigned; a ¼ Cronbach’s alpha’s for scales.