Master Thesis:
Team Process Proxies:
Appropriateness of Variance Theory Measurements
Team level analysis, Change Management
Rijksuniversiteit Groningen, Faculty of Economics and Business
January, 23, 2017
F.D. DE VRIES (s1919121) a
Word count: 15.032
a) Achterhaven 34, 9469 PR Schipborg, the Netherlands, tel. +31 0651816342, f.d.de.vries.1@student.rug.nl
Supervisors
Prof. dr. H.C. Bruns and Prof. dr. C. Reezigt
2 1. ABSTRACT
Team process is considered a crucial element in explaining team outcomes. Based on a process theory point of view, this thesis takes a critical perspective by questioning the appropriateness of variance theory in measuring team processes. In particular, I examine team process proxies that are used in the literature to function as shortcuts to represent team process in models. By performing three quantitative tests on team process proxies I show that each proxy actually represents different team processes. Inclusion of these proxies in models as general indicators of team process will therefore lead to generalizations about (the effect of) team process for team outcome that are, based on my findings, highly questionable and undesirable. Based on these findings, scholars need to start considering what type of process they measure and limit their conclusions to this particular type.
3
2. INTRODUCTION
In response to new environmental challenges and higher levels of complexity (D’Aveni, Dagnino and Smith, 2010), firms have increasingly called upon teams to manage organizational processes (Ancona, 1990; Antoni & Hertel, 2009). Scholars have argued that effective teams can be considered as important cornerstones of any successful firm (Schippers, Den Hartog & Koopan, 2007). As teams became more popular and started to be recognized as a fundamental unit of organizational structure (Drucker, 1988), considerable academic interest in this particular field began to emerge (Brannick & Prince, 1997). To be able to research team processes, many different measurement instruments of team processes (e.g. Morgan et al., 1986; Prince et al., 1989; Helmreich & Foushee, 1993; Cannon-Bowers et al., 1995) have been proposed over the years. Each instrument based upon different dimensions, factors or skills that resulted in (composite) proxies to measure different team processes (Brannick & Prince, 1997).
Parallel to the development of team process proxies, a separate discussion emerged concerning the appropriateness of using variance theory for measuring team processes (Leenders, Contractor & DeChurch, 2016). Variance theory is a mode of explanation that is state-oriented or static (Mohr, 1982; Weingart, 1997), assumes in holding constants and is thus timeless (Mohr, 1982) and its base logic tries to explain a causal relationship between independent variables and outcomes (Van de Ven, 1992) with cause and effect hypotheses (Langley, Smallman, Tsoukas & Van de Ven, 2013). In variance theory, process is not directly observed but is explained as a conduit why an independent variable exerts a causal effect on an outcome variable (Van de Ven, 1992).
The aforementioned characteristics of variance theory led process theorists to posit that methodological instruments for team processes based on variance theory are too “crude” because they lack the temporal element that process theorist consider crucial and thus not appropriate for team process measurement. According to process theory, team processes are seen as evolving phenomena where temporal progression of activities are empirically used as elements of understanding and explanation (Langley et al., 2013). By removing temporality from theoretical underpinning, it is argued that variance theory abstracts away from the temporal flow of much of the organizational life (Langley, Smallman, Tsoukas & Van de Ven, 2013). Furthermore the typical timeless propositional statements that are posited by variance theorist do not include critical elements such as; what to do at, what point in time, that actually make knowledge actionable (Sandberg & Tsoukas, 2011). Moreover process explanations that are used in variance theory to show how independent and dependent variables are causally related are considered having highly restrictive and unrealistic assumptions about the sequence and order in which events manifest themselves in organizations (Van de Ven & Huber, 1990).
4
participant constraints and lack of relevant experience (Weingart, 1997) that process theorist face when conducting process research.
Variance and process theory logic play an important role in change management as they are posited to represent different epistemologies to the study of change and development (Van de Ven & Poole, 2005). Variance theory tries to explain change in dependent variables caused by independent variables whereas process theory tries to tell a story or narrative that describes how a certain sequence of events unfolds and ultimately manifest to produce a certain outcome (Spector & Meier, 2014). It is important to corroborate or reject the aforementioned concerns of instrumental appropriateness because change management scholars need to be aware of potential fallacies before choosing to adopt such team process proxies for variance based research.
In a more general sense, given that team process is critical for explaining team outcome such as performance, it is vital for researchers to start recognizing process theory as a valuable alternative to variance theory. It is argued that process conceptualizations contribute essential knowledge to organizations and management that variance-based generalizations cannot produce (Langley et al., 2013).
A possible step towards this is by first debunking the assumption that it is possible for variance based proxies to accurately represent team process. Researchers are enticed to generalize team process and aggregate it into a single general indicator or proxy to simplify examination of models (Barrick, Stewart, Neubert & Mount, 1998) which are, following process theory logic, highly questionable and problematic. Starbuck argues that when aggregating “researchers construct homogeneity in heterogeneous phenomena” (2006, p. 143). By doing so the phenomena is inevitably simplified (Weick, 2007). The goal of this thesis is therefore to investigate the aforementioned claims of appropriateness of variance-based team process proxies and provide persuasive evidence for researchers to reconsider the applicability of variance-based instruments. To accomplish this, multiple variance-based team process proxies will be reviewed and analyzed. Moreover, a performance parameter is also included to investigate how each proxy relates to team performance. This thesis tries to answer the following research questions:
Do variance-based team process proxies differ? If so, how much do they differ and how differently do they relate to team outcome?
This thesis aims to make the following contributions to the team process literature:
1. Conducting the first multi-instrumental evaluation of team process proxies to provide a better understanding of the appropriateness and applicability of variance based measures for team process.
2. Deepen the understanding of several team process measurement instruments by exploring how they vary from eachother and how they relate to team outcome.
5
4. Setting a possible step towards resolving the confusion that is currently surrounding the team process construct by providing theoretical considerations for methodological adapation based on study findings.
6 3. THEORY
3.1. Teams
Numerous definitions of teams and groups have been used in the literature field over the years (I will use the words “team” and “group” interchangeably throughout this thesis). I follow the definition advanced by Kozlowski and Bell (2003, p. 334) which defines teams as “collectives who exist to perform organizationally relevant tasks, share one or more common goals, interact socially, exhibit task interdependencies, maintain and manage boundaries, and are embedded in an organizational context that sets boundaries, constrains the team, and influences exchanges with other units in the broader entity”.
3.2. Team process
Team process is considered a highly complex and multidimensional construct (Somech, 2006). Team process is defined as “the interaction that takes place among members” (Hackman, 1987 p. 315) and it is argued that numerous variables can reflect such intragroup processes (Cohen & Bailey, 1997). Although team processes are crucial to discover how team interaction unfold over time (Leenders, Contractor & DeChurch, 2016) and explain how a sequence of events leading from a beginning state manifest to an outcome (Spector & Meier, 2014), it is surprising how little effort is made to explicitly research how processes emerge, develop, grow or terminate over time (Langley et al., 2013).
Researchers seem too take precedence in the causal drivers of team effectiveness (Leenders, Contractor & DeChurch, 2016) and explain them in terms of covariances between independent variables and dependent variables (Van de Ven, 1992) or antecedents and consequences (Boudreau & Robey, 1999) and leave process a figuratively unopened black box or simple conduit (Mohr, 1982). This variance approach, which is by far the most common approach used in social sciences (Boudreau & Robey, 1999; Leenders, Contractor & DeChurch, 2016; Weingart, 1997; Van de Ven, 1992), is typical for most quantitative studies (Van de Ven, 1992). Although it was just mentioned how critical temporality is for understanding and explaining team processes, with the variance approach temporal sequencing is either just assumed or posited in the form of lags (Langley et al., 2013). Pairs of variables that are found significant are assumed to be adjacent steps in a process and are then used as evidence for that process (Spector & Meier, 2014).This evidence does demonstrate that a process is feasible, but it is important to note that this does not automatically establish temporal sequence as there can be numerous reasons to explain the found related pattern (Spector & Meier, 2014).
7
time are impossible to uncover (Leenders, Contractor & DeChurch, 2016) and are reduced to “comparative statistics” (Pettigrew, Woodman & Cameron, 2001, p. 697).
The variance approach assumes that temporal order runs in one direction as a recursive process (Cronin, Weingart & Todorova, 2011) which is not necessarily always the case as nonrecursive processes also exist such as flow reciprocity and feedback loops (Langley et al., 2013).
Process theory explains processes as sequences of events or activities that describe how things change and develop over time (Van de Ven & Huber, 1990; Van de Ven, 1992). Process theory allows for precise mapping of the steps of a process over time based on in-depth analysis and observations of cases (Spector & Meier, 2014). Furthermore, process theory treats the sequence of discrete events that comprise the process as holistic unit which makes the entire time span inherent to the observation (McMullen & Dimov, 2013). Each event that occurs in that time span is an indelible part or “necessary cause” (Mohr, 1982, p. 45) of the eventual explanation of the outcome that emerges (Mohr, 1982, Spector & Meier, 2014; McMullen & Dimov, 2013). In contrast with variance theory, without such event to occur it is argued that the possibility exist that the studied outcome would plausibly not occur as actually observed (McMullen & Dimov, 2013).
Now that a distinction has been made between variance and process theory and their explanative approach to process has been elucidated it is important to consider proper team process variables to use in further analyses.
Variable selection
To debunk the assumptions that it is possible for variance based proxies to accurately represent team process, variables need to be selected that purport to measure team process and represent the construct. After careful examination of the team process literature, three variables were selected that are considered generally accepted and frequently used in team process research, namely team reflexivity, team communication and team cohesion. A short description of these variables and their link to team process and team outcome is given below.
8
synergistic group interaction (Barrick et al., 1998) and a proxy to overall team functioning (Van Vianen & De Dreu, 2001). The relationship with team outcome however has been found rather ambiguous given the mixed findings that have been posited by research (Bettenhausen, 1991; Goodman, Ravlin & Schminke, 1987; Hackman, 1992).
In the next sections, the three team process proxies will be further reviewed, their respective relation with team outcome examined and the selected measurement scale presented. A comparison between each selected measurement scale will be made to examine what exactly is measured with regard to team processes based on several categories.
3.3. Team reflexivity
Team reflexivity refers to “the extent to which team members collectively reflect upon the team’s objectives, strategies and processes (West, 1996, p. 559) and is conceptually framed as a sequential and iterative process and consists of three components: reflection, planning and action / adaptation (West, 2000).
Team reflexivity is conceptualized as a sequential process that moves from reflection to planning to action and adaptation (West, 2000). This process then iterates where the action and adaptation element function as input for reflection. Reflection refers to the consideration of work-related issues and includes behaviors such as questioning, learning and analysis (West, 2000). Because reflection can be found in a variety of different types of behaviors, reflection is considered to vary in ‘depth’ of awareness of inquiry (Widmer, Schippers & West, 2009). Planning is considered to be the bridge between reflection and action or adaptation (West, 1996). Goals and ideas conceptualized in the reflection phase are articulated and ways to achieve those goals are planned (Widmer, Schippers & West, 2009) which then function as input for the action or adaptation phase. The last phase refers to the goal-directed behaviors that are directed to achieve desired changes in team objectives, strategies and processes (Widmer, Schippers & West, 2009). These changes lead to new information, which can start a new iterative cycle starting with reflection.
9
Schippers, Den Hartog and Koopman (2007) note that they are more likely highly interrelated and probably less sequential in nature.
Team reflexivity measurement instrument
In the current literature field, few measures of team reflexivity are available (Schippers, Den Hartog & Koopman, 2007). As a result, most researchers have based their results on the short questionnaire developed by Swift and West (1998). Given the apparent prevalence of this scale in team reflexivity research, I include this questionnaire as the team reflexivity measurement instrument.
Team reflexivity and team outcome
Team reflexivity has been considered an important factor for team effectiveness and performance (Schippers, West & Dawson, 2015). De Dreu (2007) for instance found a positive and direct effect of team reflexivity on team effectiveness. Team reflexivity is also considered to have a moderating impact on team interdependence and functioned as a necessary condition to foster positive effects on team effectiveness (De Dreu, 2007). Team reflexivity has been positively associated with innovation (De Dreu, Nijstad & van Knippenberg, 2008; De Dreu, 2002; Carter & West, 1998), although the investment in reflexivity was not always found worth the payoff (Bunderson & Sutcliffe, 2003). Other scholars have found similar conclusions, Moreland and McMinn’s (2010) review indicate that the relationship between reflexivity and performance can have positive consequences but often under specific conditions.
Researchers have started to incorporate contingency based thinking in their reflexivity research and suggest specific conditions where team reflexivity has the most potential (Schippers, West & Dawson, 2015). For instance, Schippers, Homan and Knippenberg (2013) have found that team reflexivity can be particularly helpful in low-performing teams as they benefited more in terms of both learning and improved final performance.
Team reflexivity has also been associated with idea exchange processes. Paulus and Yang (2000) find that, under the right conditions (i.e. contingencies), idea exchange processes can enhance innovation and creativity in organizations which in turn impact team effectiveness.
It is argued that when teams are reflecting on work processes, they may innovate and create new processes in which they can work more effectively (Schippers, West & Dawson, 2015). Such idea generating in groups has shown to be more efficient if teams have higher levels of team reflection (Paulus & Yang, 2000).
3.4. Team cohesion
10
Cohesion was first defined in the seminal work of Festinger (1950) as ‘the resultant forces which are acting on the members to stay in a group” (p. 274) but was subsequently heavily criticized (Gross & Martin, 1952). Given the process perspective of this thesis I will follow Carron (1982) and define cohesion as ‘a dynamic process which is reflected in the tendency for a group to stick together and remain united in the pursuit of its goals and objectives’ (p. 124). In the current literature field, there seem to be multiple models of cohesion with no single definition or model that is generally accepted (Cota et al., 1995). Moreover, there seem to be no agreement whether cohesion can be considered a unidimensional or multidimensional construct (Cota, Dion & Evans, 1993; Enoch & McLemore, 1967). The measuring and conceptualization inconsistencies associated with cohesiveness have problematic consequences for the literature, for instance it becomes more difficult to make sense of the literature regarding the relationship between cohesiveness and performance (Hornsey, Dwyer & Oei, 2007) which will be discussed in more detail in the section regarding team cohesion and team outcome.
It is argued that cohesion reflects synergistic interactions between team members, such as effective workload sharing, positive communication and conflict resolution (Barrick et al., 1998). Cohesive groups are associated with increased efficiency of language behavior (Mickelson & Campbell, 1975), greater use of transactive memory systems (Hollingshead, 1998, 2000; Wegner, Erber & Raymond, 1991) and greater team mental model convergence (Matthieu et al., 2000). Moreover, highly cohesive groups are found to spend more time to planning and problem solving (Shaw & Penrod, 1962), make more high-quality decisions (Janis, 1972) and work harder for the group to be successful (Goodacre, 1951; Berkowitz, 1954). Gross and Martin (1952) propose that cohesion can be thought of as “sticking-togetherness” of the group or team, or its ability to resist and overcome potentially disruptive forces. In a similar vein, Van Vianen and De Dreu (2001) describe cohesion as the extent to which people “come to stick together in groups” (cited from Muldoorn, 1955, p. 257). Team cohesion is considered to be a proxy to overall team functioning (Van Vianen & De Dreu, 2001).
It is argued that team cohesion comprises of two dimensions namely; task cohesion and social or interpersonal cohesion (Chiocchio & Essiembre, 2009). Task cohesion refers to a team member’s attraction to the group because a shared commitment to the group task exists (Brawley, Carron & Widmeyer, 1987). Social cohesion is conceptualized as a team member’s attraction to the group because of positive relationships with other group members (Zaccaro, 1991).
11
Team cohesion measurement instrument
After careful examination of the cohesion literature, at least ten formal cohesiveness measurements scales can be identified (Hornsey, Dwyer & Oei, 2007). One of these scales is a widely used instrument developed by Stokes (1983) that measures social cohesiveness. This instrument has been used in several previous studies that focus on team process and performance (e.g. Barrick et al., 1998; O’Reilly, Caldwell & Barnett, 1989; de Jong, Streukens & Ouwersloot, 2001). The scale measures three dimensions of cohesion namely; attraction to members of the group, risk taking in group and the instrumental value of the group. All three dimensions are regarded as conceptually different and predictors of social cohesiveness (Stokes, 1983).
Team cohesion and team outcome
When reviewing the team literature, there seem to be a lack of consensus regarding the direction, magnitude or even existence of a cohesion-performance relationship (e.g. Bettenhausen, 1991; Goodman, Ravlin & Schminke, 1987; Hackman, 1992). There are researchers who find a positive relationship and identify cohesion as a factor that exert influence on group performance (Stogdill, 1972; Wech, Mossholder, Steel & Bennett, 1998). A number of meta-analyses (Gully, Devine & Whitney, 1995; Mullen & Copper, 1994) also showed a positive relationship between cohesion and performance, although sometimes moderated by other construct. Beal, Cohen, Burke and McLendon (2003) posit that cohesive groups are more efficiently using the groups’ resources because they know the group members better and are better motivated to complete tasks successfully and thus more productive. They further argue that, with regard to team outcome, effectiveness measures may be less sensitive to the aforementioned aspects whereas efficiency measures are better designed to detect these sorts of processes. Other studies however reported more mixed results (e.g. Forsyth, 1990), Chang and Bordia (2001) found only a temporal effect whereas others (Gladstein, 1984; Steiner, 1972) found no evidence at all for the existence of a relationship between cohesion and performance. It is argued that in general, there tend to be more support for a positive relationship between cohesion and performance (Swamidass, 2003; Chiocchio & Essiembre, 2009).
12 3.5. Team communication
Communication has been regarded as the heart of group behavior (Shaw, 1981) and is considered to be the essence of social systems (Katz & Kahn, 1978). It is argued that group communication is multidimensional and is broken down into communication frequency (Daft & Lengel, 1984; Ancona & Caldwell, 1992) and communication informality (Katz & Kahn, 1978). Communication frequency envelops the way of communication (e.g. face-to-face, telephone, mail) and the amount of team member interaction (Shaw, 1981). Communication informality concerns the extent to which teams favor less formal communication channels (e.g. spontaneous conversations), over more formal channels (e.g. written communication) (Smith et al., 1994). Others have described communication in delineations such as internal versus external communication, formal versus informal communication and written versus oral communication (Pinto & Pinto, 1990). As can be observed, there seems to be some overlap between communication classifications of this particular construct.
With regard to empirical findings of communication, it has been found that communication partners are more willing to share knowledge when teams communicate in an agreeable manner (de Vries, van den Hooff & de Ridder, 2006). Moreover, different communication styles are associated as important predictors of attitudes. For example, whether bank employees create customer satisfaction (Wong & Tjosvold, 1995) or whether somebody is seen as a leader (Awamleh & Gardner, 1999).
Communication is regarded as an important element for building and maintaining a productive interface between functional units (Carroad & Carroad, 1982; Dutton & Walton, 1966; Link & Zmud, 1986). Some authors consider communication to be the vehicle of new product development because team members use communication to share information on markets, competitors and technologies to solve product design problems and create technical solutions (Leenders, van Engelen & Kratzer, 2003). Communication as a process has been found to be pervasive to all activities of managers, as communication managerial skills (e.g. interacting and allocating) are operationalized only by communication activities (Pinto & Pinto, 1990).
Team communication measurement instrument
For team communication, the scale of Temkin-Greener, Gross, Kunitz & Mukamel (2004) was used. Although there are numerous team communication scales available, there were certain methodological restrictions concerning the length of each instrument. Good reliability and validity was reported in the instrument testing of this particular scale and given it was the proper length, it was included.
Team communication and team outcome
13
Moreover, Ouchi (1980) found that highly communicative teams have lower communication and coordination costs and are considered more flexible and efficient. Furthermore, communication factors have been considered to affect important outcomes in organizations (e.g., Falcione, 1974; DeWine, 2001; Downs and Hazen, 1977). Tushman (1977) found that flexible communication patterns are associated with high performing groups under highly uncertain tasks.
There are also scholars who found a negative relationship between communication and team performance. For instance, Ancona and Caldwell (1992) found a negative relationship between communication frequency and performance. They argue that communication frequency in fact can indicate team conflict, which has a negative effect on team performance. This finding was corroborated by Smith and colleagues (1994) who found that communication frequency indeed indicated team conflict and thus decreased team decision-making speed and indirectly performance (Eisenhardt, 1989).
3.6. Comparison
The following comparison is based on a combination of information gathered from the articles’ methodological model and logic, their respective items and my own interpretation of the instrument. Each instrument is broken down into six categories that represent methodological characteristics and facets of team process. The goal of this comparison is to uncover how each proxy relates to the six categories and in doing so providing a deeper and better understanding of what each proxy actually tries to measure with regard to the team process construct. The comparative table is shown below.
Table 1
Iterative Scale Comparison Between Three Team Process Proxies Categories Team reflexivitya Team cohesionb Team communicationc Reflective or
formative items
Both Both Reflective
Dimensions Unidimensional Multidimensional Multidimensional Internal or external
process
Internal and external process
Internal process or group psychosocial trait
Internal process
Type of response (affective or cognitive)
Cognitive Affective Cognitive
Latent variables Team reflection Goal reflection Process reflection
Trust in team / Risk Team attraction Group value
Information accuracy Communication effectiveness
Dynamic or State State State State
a Based on Swift and West (1998) b Based on Stokes (1983)
14
With respect to scale development, reflective items or indicators are by far the most commonly used to measure latent variables (Law & Wong, 1999). Reflective indicators can be seen as functions of the latent variables, whereby a change in the latent variable is also reflected in a change in the observable indicator (Diamantopoulos & Siguaw, 2006). Formative indicators are formulated in the other way, as changes in the formative indicators determine changes in the value of the latent variable (Jarvis, Mackenzie & Podsakoff, 2003). In other words, reflective indicators can be seen as interchangeable, and the direction of causality is from construct to measure whereas formative indicators are not interchangeable and the direction of causality is from measure to construct. After inspecting the instruments, which are all measuring latent variables using supposedly reflective indicators, the team reflexivity and team cohesion scales seem to be somewhat inconsistent as both reflective and formative indicators seem to be present. For instance, when comparing indicator 1 (‘The team often reviews its objectives’) and 7 (‘Team strategies are rarely changed’, Reversed) of the team reflexivity scale, it is perfectly reasonable for teams to often review their objectives but not have to change team strategies. This lack of interchangeability indicate the presence of formative indicators. For the team cohesion scale, when comparing indicator 6 (‘I like the team that I am in’) and 7 (‘My team should meet often’) that are part of the latent variable ‘group value’, the same interchangeability problem is present as in the previous example. This combination of reflective and formative indicators is problematic because it is argued to be prone to Type I and Type II errors and should be avoided (Jarvis, Mackenzie & Podsakoff, 2003; Diamantopoulos & Siguaw, 2006).
The team reflexivity scale is the only unidimensional scale used as it represents one dimension; reflection (Schippers, Den Hartog & Koopman, 2007). Team cohesion represents three dimension; Risk, or trust in team, team attraction and value of the group (Stokes 1983). Team communication represents two dimensions; Information accuracy and communication effectiveness.
Both team reflexivity and team communication have internal and external elements with regard to team processes (Cohen & Bailey, 1997; Widmer, Schippers & West, 2009). The team communication scale however does not have any indicator that reflects inter-team interaction and thus is categorized as an internal process only. Team cohesion is inherently an internal process as it is centered on the perceived attractiveness and value of the team by a team member (Brawley, Carron & Widmeyer, 1987). Team cohesion can also be considered as a group psychosocial trait instead of an internal process (Cohen & Bailey, 1997).
15
cohesion however taps into a more affective dimension and is associated with more ‘emotional response’ because they are more subjectively framed and could be considered more relationship-oriented.
The latent variables represent the dimension(s) of each proxy and gives further insight into what facets of team process are considered and included in each proxy. Each proxy clearly emphasizes different processes in teams which raises concerns of the applicability of such proxies in models aimed to generalize the effect to ‘one’ team process where in fact multiple team processes are in effect.
With respect to the dynamic nature of team processes (McGrath, Arrow & Berdahl. 2000), which incorporate the temporal nature of team interactions (Langley et al. 2013), the proxies are all measuring a fixed state in one point in time. The proxies do not consider any form of dynamic interaction or temporal aspects within the team.
16
4. METHODOLOGY
4.1. Study context
I sampled work teams that were predominantly full-time employees from four companies in the Netherlands. The first company was a high tech agricultural processing organization where teams researched and produced biologically-modified seeds for international and national businesses and direct consumers. The other three organizations were situated in the telecom and electronics industry, energy (gas) industry and furniture design industry respectively. These sampled teams were located throughout the organizations (e.g. R&D, human resources, IT) which increases external validity of the possible relationships found between variables (Hu & Liden, 2015).
4.2. Data and procedure
Data
The two main sources of data used in this thesis were a paper and pencil survey distributed to team leaders of a high tech agricultural processing organization and an online survey that was distributed through the intranet of the three aforementioned organizations. Team leaders not present at the time the pencil and paper survey was conducted were send the online survey instead. The online survey was also directed at team leaders within each organization. The final sample included 73 teams and was deemed adequate for a reliable regression analysis because it exceeded the 15 participants per predictor threshold (Steven’s, 1996).
The average team size comprised of 5.89 members (SD = 3.21). The average team tenure was 4.03 years (SD = 2.88) and average team task interdependence was 3.98 (SD = 0.68). The response rate of the paper and pencil survey was 79%.
With respect to missing values of team size, tenure and task interdependence the mean value was taken instead. Surveys with missing values on the independent and dependent variables were deleted and not used in further analyses.
Procedure survey
17
Procedure validating team process proxies
In order to quantitatively investigate the three variance based proxies of team process for appropriateness, I conduct three tests to provide insight how each proxy represent and measures team processes. First, a raw (without data correction) explanatory factor analysis will be used to examine the explained variance for different factor settings. Second, a correlation matrix will show whether or not the proxies’ correlations are significantly different from each other. This will be done by correcting the raw data to remove cross loading and inappropriate loadings which result in clean factors of which each cluster of items represent the appropriate proxy. These clean factors will then be used to compute respective Cronbach Alpha’s and construct mean factor scores, which will be the input for the correlation matrix. Third, beta scores will be compared between each proxy and a performance construct to examine if the proxies significantly differ in explaining the performance construct in direction and magnitude. This particular step starts by using the same data correction procedure for the dependent variable which produces a clean dependent variable factor. Then a hierarchical regression will be produced for the performance construct by sequentially adding control variables and then the independent variables. The three tests are presented in the results section.
4.3. Measures
The sampled teams were located in the Netherlands, and as the team members were natively Dutch speaking, I presented all the material in Dutch. Since the original measurement scales used were in English, checks were applied to test the appropriateness of the translation of the scales. First, the scales were translated by myself and a university colleagues separately. Following the procedure of Bullock and Svyantek (1985) the inconsistencies in the translations were discussed and reexamined until agreement was found. The translated scales were then presented to an unbiased third university colleague that was asked to translate the scales back to English. These scales were then compared to the original scales, differences were discussed and appropriate corrections were made to the translated scales.
Furthermore, adjustments have been made based on the comments of teams and supervisors that were asked to give feedback on the first draft of the questionnaire. The items in the instrument scales were slightly rephrased and some minor changes were made to be more appropriate and relevant for the study context. Also, some items have been left out as they were not applicable in the study context. The final questionnaire consisted of items that were considered understood and seen as relevant and unambiguous. The used constructs and related measurement scales are presented below.
Team reflexivity
18
will talk about it with other team members’ and ‘After certain activities are completed, we evaluate matters’. The items were rated on a five-point scale (1 = strongly disagree, 5 = strongly agree).
Team cohesion
Team cohesion was measured using a 7 item scale made by Stokes (1983). Examples of items are: ‘I feel included in group activities’ and ‘I like the team that I am in’. The items were rated on a five-point scale (1 = strongly disagree, 5 = strongly agree).
Team communication
Team communication was measuring used a 5 item scale made by Temkin-Greener, Gross, Kunitz and Mukamel (2004). Examples of items are: ‘There is effective communication between team members about their objective’ and ‘Information passed between team members is accurate’. The items were rated on a five-point scale (1 = strongly disagree, 5 = strongly agree).
Team effectiveness
Team effectiveness was measured using a 5 item scale made by Campion, Papper and Medsker (1996). The scale contains both efficiency and effectively items. Examples of items are ‘The team’s productivity can be considered high’ and ‘The work done of my team is of high quality’. It should be noted that team effectiveness is self-reported and could therefore be prone to bias. The items were rated on a five-point scale (1 = strongly disagree, 5 = strongly agree).
4.4. Control variables
Team Size
Team size was controlled because it has been found that team size can positively impact team performance (Haleblian & Finkelstein, 1993). Haleblian and Finkelstein (1993) argue and find that an increase in team size represents relatively more cognitive resources that can be used to reach higher performance. Furthermore, Group size is associated with the potential for more heterogeneity and the assumption that it may influence group outcomes (Mohammed & Angell, 2004). There are researchers who posit that team size can sometimes negatively relate to team outcomes because of process loss (e.g. Paulus et al, 2002). Given the abovementioned effects, it is important to include team size as a control variable.
Team tenure
19
Team task interdependence
20 5. RESULTS
In order to test each team process proxy, I first conducted a series of exploratory factor analyses, using a varimax factor solution in SPSS. I used a varimax factor solution because I expected the extracted factors to be independent rather than correlated and thus an orthogonal factor solution was deemed more adequate.
Considering the assumption that each proxy measures the same team process, a single factor solution (i.e. team process) should logically emerge and will explain the most variance. To test this assumption, I ran three free explanatory factor analyses that were extracted based on a fixed number of factors and then analyzed the cumulative variance each factorial solution explained. The results of this test are shown below.
Table 2
Total Variance Explained Per Factor Rotation
Component
Initial Eigenvalues
Extraction Sums of Squared Loadings
Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance Cumulative % EFA 1a 1 6,648 31,657 31,657 6,648 31,657 31,657 EFA 2 1 6,648 31,657 31,657 6,648 31,657 31,657 6,519 31,041 31,041 2 2,380 11,334 42,991 2,380 11,334 42,991 2,509 11,950 42,991 EFA 3 1 6,648 31,657 31,657 6,648 31,657 31,657 6,017 28,652 28,652 2 2,380 11,334 42,991 2,380 11,334 42,991 2,379 11,328 39,980 3 1,718 8,181 51,172 1,718 8,181 51,172 2,350 11,192 51,172 Extraction Method: Principal Component Analysis.
a EFA 1 could not provide rotation sums of squared loadings as it exceeded 25 rotations to compute.
As can be observed in table 2 the cumulative variance explained, which is shown in bold, is substantially higher with three fixed factors (51.17%) than one fixed factor (31.66%). Even a two factor fixed solution (42.99%) explains more variance than the one fixed factor solution. These findings thus contradict the assumptions that each proxy represents the same team process and indicates that they are clearly measuring other team processes as well.
21
because you can only truly compare each proxy for intercorrelations if they are fully independent of each other and can be compared as separate and distinct instruments. Several reliability and internal consistency criteria were used to accomplish the aforementioned homogeneity and items that did not meet these criteria were deleted. The first criterion used was a Cronbach’s alpha of > .60 as a first indicator of scale reliability. As a second criterion, factor loadings should be roughly >0.65 considering my sample size of > 70 (Hair, Anderson, Tatham & Black, 1998). The third criterion used requires the factor loadings of an item to differ > .20 (Den Hartog, Van Muijen & Koopman 1997). The fourth criterion used was a > .4 extraction communality. Extraction communality refers to the variance accounted for by that item in the total factor solution, small values thus indicate a poor fit with the solution and should be dropped. Twelve items did not meet criteria and were deleted from further analyses. Table 3 shows the factor loadings of each items on its respective factor.
Table 3
Loadings on the Itemsa of Reflexivity, Cohesion and Communication
Respectively
Factor
1 2 3
1. Discuss if team is working effectively ,769 2. Reviews if team gets the job done ,839 3. Discussed the methods used by team ,887 4. Discuss communication of information ,709 5. Reviewing objectives of team ,649
6. Dissuade team dissolvent of member ,858
7. How often the group should meet ,834
8.Team members understand each other ,830
9.Effective communication about objective ,807 Note: N = 73. PCA with varimax rotation.
a. Short version of the items are given. Full text items can be found in the Appendix.
Next, these clean factors will be used to construct mean factor scores for team reflexivity, team cohesion and team communication respectively, which will be input for the correlation matrix, subsequent analyses and computation of the Cronbach alpha’s. Table 4, which is provided below, shows the means, standard deviations, Cronbach alpha’s, and intercorrelations of all independent variables and the control variables. As can be observed, each variables’ Cronbach alpha meets the reliability threshold of 0.6.
22
the three proxies reach intercorrelations levels that validate the aforementioned assumptions. These findings are in line with the earlier findings and provide contradicting evidence that each proxy supposedly represents the same team process construct.
I conducted one more analysis to investigate the assumption that each proxy measures the same team process. To prove or dispute this assumption, I used hierarchical regression to compare beta scores between each proxy and a performance construct. If each proxy measures the same team process, beta scores should logically be similar in direction and magnitude when explaining the performance construct. To conduct this analysis, I started by using the same data correction procedure for the dependent variable which produced a clean dependent variable factor ( = .822) after deleting 1 item. Next, I consider the appropriateness of the chosen control variables for inclusion in the hierarchical regression. Table 4 shows that the relationship between team size and team task interdependence was substantial and significant (r = .358, p < .01), team size and team reflexivity also had a positive significant relationship (r = .293, p < .05). Team tenure was significantly correlated to team reflexivity (r = .233, p < 0.05) and team task interdependence was found to have a significant relationship with communication (r = .3, p < 0.01). Abovementioned findings indicate that the control variables are potent and were therefore all included in the hierarchical regression. For the regression analysis, team size, team tenure and team task interdependence were entered as the first step and team reflexivity, team cohesion and team communication as the second step. Results of the regression analysis are presented in table 5.
Table 4
Means, Standard Deviations and Intercorrelations
M SD 1 2 3 4 5 1. Team size 5.89 3.21 2. Team tenure 4.03 2.88 0.075 3. Team task interdependence 3.98 0.68 0.78 0.358** -0.094 4. Team reflexivity 3.47 0.83 0.86 0.293* 0.233* 0.364 5. Team communication 3.87 0.89 0.65 0.063 -0.064 0.3** 0.462** 6. Team cohesion 3.24 1.04 0.63 0.033 0.041 -0.099 0.205 0.168 Note: N = 73 teams.
23 Table 5
Hierarchical Regression With Dependent Variable Team Effectiveness
Model 1 Model 2
Variable β t Sig. β t Sig.
Control Variables Team tenure ,055 ,508 ,613 ,035 ,342 ,734 Team size ,012 ,104 ,917 ,022 ,210 ,835 Team task interdependence ,451 3,896 ,000** ,259 2,271 ,026* Main effects Team reflexivity ,145 1,173 ,245 Team communication ,384 3,400 ,001** Team cohesion -,192 -1,903 ,061 R2 .21 5.97 (p < 0.001) .39 6.7 (p < 0.00) F for change in R2 (sig.)
*p < .05. **p < .01.
Beta coefficients for the three proxies were respectively team reflexivity, β = .145, t = 1.173, p < .245, n.s.; team communication, β = .384, t = 3.400, p < .001; and team cohesion, β = -.192, t = -1.903,
p < .061, n.s.. Findings from this regression analysis thus show that there is little similarity in direction
24
6. DISCUSSION AND CONCLUSION
The results of this thesis provide convincing evidence that the team process measures, forwarded by researchers to function as proxies for team process, are clearly representing more than one team process. The tests of comparing cumulative variance explained, the intercorrelations between each proxy and their respective relationship with performance all clearly indicate that each proxy is distinct which rules out the possibility that each proxy represents the identical underlying construct. Researchers who include team process as a variable for various larger models (e.g. input-process-output models) or empirical research often seek to simplify examination and thus develop hypotheses that adopt proxies as a general indicator of team process (Barrick et al., 1998). Selecting team process proxies for empirical research that are actually measuring several different team processes but purport to represent team process, as though only one team process exists, are inherently problematic and result in inaccurate results and conclusions which hurt the literature field.
Conceptually, team researchers seem to slowly converge on the idea imbedded into process theory that teams and their processes are adaptive, dynamic systems (McGrath, Arrow & Berdahl. 2000) and that such interactions change teams and team members in a way that is too complex for simple cause and effect perspectives to capture (Ilgen, Hollenbeck, Johnson & Jundt, 2005; Van de Ven, 1992). Empirically however these notions are more often than not suppressed, not considered or deemphasized in significance (Dawson, 2014) because of lack of experience or feasibility (Weingart, 1997). The results of this thesis clearly demonstrate that the proxies used in this research are not measuring one and the same process and therefore making generalized conclusions based on such proxies is questionable and inappropriate.
It is important to understand and respect the obvious multidimensionality of team process and realize that the selection of one specific variable does not fully capture the complexity of team process and its relation to team outcome (Somech, 2006). The comparative table that was presented in the theory section justly represents the actual narrow representation of team process that is measured by each proxy of the team process construct. Although it is of iterative nature, it demonstrates a delineation into several categories that shows the differences of each proxy. The identified differences in latent variables that were used for grounding the instrument into a model and theory is one of the differences that provide reasoning and logic behind the results of the qualitative testing. These results (i.e. cumulative variance explained and the three factor solution) clearly indicate that each proxy is distinct and measures separate variables or dimensions (although not in the same magnitude, direction or causality) that can be considered separate processes in teams.
25
A possible explanation for the insignificant relationship between these two variables and team effectiveness is the fact that variance theory based instruments lack the element of temporality (Sandberg & Tsoukas, 2011). As already said mentioned multiple times, process is inextricably linked to time because it represent a temporal sequence of activities, states or events (Langley et al., 2013). An important element of temporality involved in process is the so called equilibration period that is defined as ‘the amount of time it takes for an antecedent to affect a consequent and reach a steady state’ (Spector & Meier, p. 1111). Each process variable will theoretically have different equilibration periods and affect a consequent (i.e. team outcome) will take varying amount of time based on the length of the equilibration period.
It is very much plausible that the insignificant findings were product of the timeless research design of variance theory based measurement which observed an antecedent that in that particular time lag or measuring point (Langley et al., 2013) was still a steady state ‘in becoming’ or ‘emerging’ and thus did not sufficiently affect a consequence to reach a level of significance. In other words, because the variance theory based measurement does not take into account the temporal interval that was needed for the antecedent to ‘grow’ and ‘develop’ and ‘emerge’ (Van de Ven & Huber, 1990) into a significant enough steady state, it found insignificance. When temporal elements were taken into account and proper equilibration periods were followed significance could have been demonstrated.
With respect to the equilibration periods of team cohesion and team reflexivity it could be argued that, based on theory between the independent variables and team outcome, they are relatively long compared with team communication. Team cohesion has been found to takes considerable time before team members develop some kind of attraction to other team members (Chiocchio & Essiembre, 2009). Team reflexivity is associated with learning, divertive exploration, analysis and coming to terms over time with new awareness (West, 1996, 2000) which are all rather lengthy processes. The entire process is even divided in several phases (West, 1996) which emphasizes a relatively long sequence of events. Team communication on the other hand seem to be a pervasive element to all activities of managers and team members, which can be identified as series of short sequences of events that need short time to manifest. It thus can be assumed that each individual sequence of communication has a short equilibration period and is therefore less susceptible for the abovementioned concern of the amount of time it takes to manifest into a steady state than team reflexivity and team cohesion.
With regard to the research question, this thesis has theoretically and quantitatively shown that each proxy in fact measures different team processes and does not represent the same underlying process.
6.1. Theoretical implications
26
From a process theory point of view, researchers should take note of the lack of temporal references in these proxy measurements which calls in question the appropriateness of these proxies. As been illustrated in the theory section, team processes can be considered a sequence of events that emerge over time (Boudreau & Robey, 1999; Markus & Robey 1988) and represent dynamic interactions between team members (Somech, 2006). By not including any temporal references in the proxies, these measurements seem to abstract away from the essence of team interaction and temporal flow that runs through organizations and teams (Langley et al., 2013). Based on these arguments the usefulness of these proxies seems highly limited. Van de Ven and Huber (1990) reached to similar conclusions as they also found timeless variance based studies to be highly restrictive and unrealistic in their assumptions about the dynamics of team processes and event manifestation.
6.2. Managerial implications
The gap that exist between theory and practice has been a particular concern in the applied social sciences and seem to be persistent (Lindblom & Cohen, 1979; Lupton, 1983; Sandberg & Tsoukas, 2011). Managers are increasingly dissatisfied with the capacity of management theory to be relevant for practice (Ghoshal, 2005; Starbuck, 2006) and that knowledge produced by research is too distant for managers to be useful (Ghoshal, 2005; Van de Ven & Johnson, 2006). Researchers are argued to contribute to this problem by conceptualizing practice as an atemporal space (Sandberg & Tsoukas, 2011). Although variance based research is considered to be rather pragmatic (Cornelissen, 2016), the negligence of temporality in most of variance measurement instruments seem to contribute to this problem. The same atemporal conceptualization was also identified in the measures of proxies used in this thesis. Because the gross of studies is variance based (Weingart, 1997; Van de Ven, 1992), managers are encouraged to also consider process theory oriented works for managerial application. Process theory studies better reflect the temporal flow of practice such as urgencies, practical necessities and uncertainties that managers find themselves in all so often. Process theory studies in turn provide in-depth and rich answers of what to do, at what point in time (Sandberg & Tsoukas, 2011) that can be of particular significance for managers in similar situations.
6.3. Limitations and suggestions for future research
A possible limitation of this thesis is the extensive scale adaptation that had to be done to operationalize the scales. Although procedures, protocols and checks were in place to minimize possible translation mistakes or adaption failures, the resulting translated scales could still be interpreted differently than the original which should be considered before interpreting the results.
27
Future research is encouraged to further investigate the team process proxies and their use in social sciences. A possible avenue for future research could be to develop a parallel approach that incorporates both variance and process theory logic that for instance target the same team process (e.g. Van de Ven & Poole, 2005). In doing so, more conclusive evidence could be obtained how each approach measures the type of team process, what their communalities and differences are, how they relate to team outcome and if they have any complementarity that could have been overlooked. Indeed, Mohr (1982) suggested that process and variance theory can be mutually informative. Moreover, Maynard (2008) found that different process constructs are better measured by different types of measures (e.g. retrospective or observations), that can either be process or variance theory based. Cohesion for instance is argued to be hard to externally asses which makes self-assessment a better alternative (Maynard, 2008).
Researchers are encouraged to replicate this particular experiment using different team process proxies (e.g. team interaction, team coordination or strategic planning, Hackman, 1987) to solidify and corroborate the current research findings. Researchers should also consider using different or multiple team outcome measures and examine how the various proxies relate to different team outcome measures.
6.4. Conclusion
Team processes are multidimensional and complex constructs that emerge, develop, grow, or terminate over time (Langley et al., 2013) and need to be recognized as such. This research showed that the practice to include timeless proxies of team process lead to generalizations of the effect of team processes on team outcome which are highly questionable and problematic. Scholars need to start considering what type of team process they adopt in their research and should limit their conclusions to this particular type.
With respect to the epistemological differences for the study of change, this thesis does not in any way posit that variance theory is inappropriate for the study of change. On the contrary, when taking a more pluralist point of view, it is believed that variance and process theory can be used as complimentary motors of explanation (Van de Ven & Poole, 2005) to describe change in a more holistic and complete manner than either of the theories are able on their own (e.g. Saberwhal & Robey, 1995). This view in turn helps develop a more objective scientific knowledge of change management as multiple plausible models of realities are being compared and used in unison (Campbell, 1988).
28
Scholars need to realize that there is no such thing as ‘one overarching team process’ that can be encapsulated into a proxy and instead look closer at what particular team process they are actually examining. Recognizing relevant team process types for conceptualization of models and subsequent conclusions is a first step that will inevitably lead to an enrichment of the process related literature field.
6.5. Acknowledgements
29 7. REFERENCES
Ancona, D. G. (1990). Outward bound: strategic for team survival in an organization. Academy of Management journal, 33(2), 334-365.
Ancona, D. G., & Caldwell, D. F. (1992). Bridging the boundary: External activity and performance in organizational teams. Administrative science quarterly, 634-665.
Antoni, C., & Hertel, G. (2009). Team processes, their antecedents and consequences: Implications
for different types of teamwork. European Journal of Work and Organizational Psychology, 18(3),
253-266.
Awamleh, R., & Gardner, W. L. (1999). Perceptions of leader charisma and effectiveness: The effects of vision content, delivery, and organizational performance. The Leadership Quarterly, 10(3), 345-373.
Barrick, M. R., Stewart, G. L., Neubert, M. J., & Mount, M. K. (1998). Relating member ability and
personality to work-team processes and team effectiveness. Journal of applied psychology, 83(3), 377. Barry, B., & Stewart, G. L. (1997). Composition, process, and performance in self-managed groups: the role of personality. Journal of Applied psychology, 82(1), 62.
Beal, D. J., Cohen, R. R., Burke, M. J., & McLendon, C. L. (2003). Cohesion and performance in groups: a meta-analytic clarification of construct relations.
Berkowitz, L. (1954). Group standards, cohesiveness, and productivity. Human Relations, 7(4), 509-519.
Bettenhausen, K. L. (1991). Five years of groups research: What we have learned and what needs to be addressed. Journal of management, 17(2), 345-381.
Bollen, K. A., & Hoyle, R. H. (1990). Perceived cohesion: A conceptual and empirical examination. Social forces, 479-504.
Bostrom, R. N. (1970). Affective, cognitive, and behavioral dimensions of communicative attitudes. Journal of Communication, 20(4), 359.
Boudreau, M. C., & Robey, D. (1999). Organizational transition to enterprise resource planning systems: theoretical choices for process research. In Proceedings of the 20th international conference on Information Systems (pp. 291-299). Association for Information Systems.
Brannick, M. T., & Prince, C. (1997). An overview of team performance measurement. Team
performance assessment and measurement. Theory, methods, and applications, 3.
Brawley, L. R., Carron, A. V., & Widmeyer, W. N. (1987). Assessing the cohesion of teams: Validity of the Group Environment Questionnaire. Journal of Sport Psychology, 9(3), 275-294.
Buckner, J. C. (1988). The development of an instrument to measure neighborhood cohesion. American journal of community psychology, 16(6), 771-791.
30
Bunderson, J. S., & Sutcliffe, K. M. (2003). Management team learning orientation and business
unit performance. Journal of Applied Psychology, 88(3), 552.
Campbell, D. T. (1988). Methodology and epistemology for social sciences: Selected papers. University of Chicago Press.
Campion, M. A., Papper, E. M., & Medsker, G. J. (1996). Relations between work team characteristics and effectiveness: A replication and extension. Personnel psychology, 49(2), 429-452.
Cannon-Bowers, J. A., Prince, C., Salas, E., Owens, J. M., Morgan Jr, B. B., & Gonos, G. H. (1989).
Determining aircrew coordination training effectiveness. In Proceedings of the 11th Annual Meeting of
the Interservice/Industry Training Systems Conference. Ft Worth, TX (pp. 128-36).
Cannon-Bowers, J. A., Tannenbaum, S. I., Salas, E., & Volpe, C. E. (1995). Defining competencies and establishing team training requirements. Team effectiveness and decision making in organizations, 333, 380.
Carroad, P. A., & Carroad, C. A. (1982). Strategic interfacing of R&D and marketing. Research Management, 25(1), 28-33.
Carron, A. V. (1982). Cohesiveness in Sport Groups: Interpretations ani Considerations. Journal of Sport psychology, 4(2), 123-138.
Carron, A. V., & Brawley, L. R. (2000). Cohesion conceptual and measurement issues. Small Group Research, 31(1), 89-106.
Carter, S. M., & West, M. A. (1998). Reflexivity, effectiveness, and mental health in BBC-TV
production teams. Small group research, 29(5), 583-601.
Chang, A., & Bordia, P. (2001). A multidimensional approach to the group cohesion-group performance relationship. Small Group Research, 32(4), 379-405.
Chiocchio, F., & Essiembre, H. (2009). Cohesion and performance: A meta-analytic review of disparities between project teams, production teams, and service teams. Small group research.
Cohen, S. G., & 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.
Cornelissen, J. P. (2016). Preserving theoretical divergence in management research: Why the explanatory potential of qualitative research should be harnessed rather than suppressed. Journal of Management Studies.
Cota, A. A., Evans, C. R., Dion, K. L., Kilik, L., & Longman, R. S. (1995). The structure of group
cohesion. Personality and Social Psychology Bulletin, 21(6), 572-580.
Cota, A. A., Dion, K. L., & Evans, C. R. (1993). A reexamination of the structure of the Gross Cohesiveness Scale. Educational and Psychological Measurement, 53(2), 499-506.
Cronin, M. A., Weingart, L. R., & Todorova, G. (2011). Dynamics in groups: Are we there yet?. Academy of Management Annals, 5(1), 571-612.
31
Daft, R. L., & Lengel, R. H. (1984). Information richness: a new approach to manager information processing and organisational design. Research in Organisational Behaviour.
Dawson, P. (2014). Reflections: on time, temporality and change in organizations. Journal of Change Management, 14(3), 285-308.
Dreu, C. K. D. (2002). Team innovation and team effectiveness: The importance of minority dissent
and reflexivity. European Journal of Work and Organizational Psychology, 11(3), 285-298.
De Dreu, C. K. (2007). Cooperative outcome interdependence, task reflexivity, and team
effectiveness: a motivated information processing perspective. Journal of Applied Psychology, 92(3),
628.
De Dreu, C. K., Nijstad, B. A., & van Knippenberg, D. (2008). Motivated information processing in
group judgment and decision making. Personality and Social Psychology Review, 12(1), 22-49.
de Jong, A., de Ruyter, K., Streukens, S., & Ouwersloot, H. (2001). Perceived uncertainty in self-managed service teams: an empirical assessment. International Journal of Service Industry
Management, 12(2), 158-183.
Den Hartog, D. N., Van Muijen, J. J., & Koopman, P. L. (1997). Organisatiecultuur en leiderschap (Organisational culture and leadership). Gedrag en Organisatie, 10, 121-133.
DeWine, S. (2001). The consultant's craft. Boston, MA: Bedford St. Martins.
De Vries, R. E., Van den Hooff, B., & de Ridder, J. A. (2006). Explaining knowledge sharing: The role of team communication styles, job satisfaction, and performance beliefs. Communication research, 33(2), 115-135.
Diamantopoulos, A., & Siguaw, J. A. (2006). Formative versus reflective indicators in organizational measure development: A comparison and empirical illustration. British Journal of Management, 17(4), 263-282.
Donnellon, A. (1993). Crossfunctional teams in product development: Accomodating the structure to the process. Journal of Product Innovation Management, 10(5), 377-392.
Downs, C. W., & Hazen, M. D. (1977). A factor analytic study of communication satisfaction. The Journal of Business Communication (1973), 14(3), 63-73.
Drucker, P. (1988). The Coming of the New Organization. Harvard Business Review, 66, 45-53.
Dutton, J., & Walton, R. (1966). Interdepartmental conflict and cooperation: Two contrasting studies. Human Organization, 25(3), 207-220.
Eisenhardt, K. M. (1989). Making fast strategic decisions in high-velocity environments. Academy of Management journal, 32(3), 543-576.
Enoch, J. R., & McLemore, S. D. (1967). On the meaning of group cohesion. The Southwestern Social Science Quarterly, 174-182.
Falcione, R. L. (1974). The factor structure of source credibility scales for immediate superiors in the organizational context. Communication Studies, 25(1), 63-66.
32
Forsyth, D. R. (1990). Group Dynamics. Brooks. Cole, Pacific Grove, CA.
Ghoshal, S. (2005). Bad management theories are destroying good management practices. Academy of Management learning & education, 4(1), 75-91.
Gladstein, D. L. (1984). Groups in context: A model of task group effectiveness. Administrative science quarterly, 499-517.
Goodacre, D. M. (1951). The use of a sociometric test as a predictor of combat unit effectiveness. Sociometry, 14(2/3), 148-152.
Goodman, P. S., Ravlin, E. C., & Schminke, M. (1987). Understanding Groups in Organizations. Research in Organizational Behavior, 9, 121-173.
Gross, N., & Martin, W. E. (1952). On group cohesiveness. American Journal of Sociology, 57(6), 546-564.
Gully, S. M., Devine, D. J., & Whitney, D. J. (1995). A meta-analysis of cohesion and performance effects of level of analysis and task interdependence. Small group research, 26(4), 497-520.
Hackman, J. R. (1987). The design of work teams. Inj. w. lorsch (ed.), Handbook of organizational behavior (pp. 315-342).
Hackman, J. R. (1992). Group influences on individuals in organizations. Consulting Psychologists Press.
Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (1998). Multivariate data analysis . Uppersaddle River. Multivariate Data Analysis (5th ed) Upper Saddle River.
Haleblian, J., & Finkelstein, S. (1993). Top management team size, CEO dominance, and firm performance: The moderating roles of environmental turbulence and discretion. Academy of management journal, 36(4), 844-863.
Helmreich, R. L., & Foushee, H. C. (1993). Why crew resource management? Empirical and
theoretical bases of human factors training in aviation. Academic Press.
Hollingshead, A. B. (1998). Communication, learning, and retrieval in transactive memory systems. Journal of experimental social psychology, 34(5), 423-442.
Hollingshead, A. B. (2000). Perceptions of expertise and transactive memory in work relationships. Group Processes & Intergroup Relations, 3(3), 257-267.
Hoogstraten, J., & Vorst, H. C. (1978). Group cohesion, task performance, and the experimenter expectancy effect. Human Relations, 31(11), 939-956.
Hornsey, M. J., Dwyer, L., & Oei, T. P. (2007). Beyond Cohesiveness Reconceptualizing the Link Between Group Processes and Outcomes in Group Psychotherapy. Small Group Research, 38(5), 567-592.
Hu, J., & Liden, R. C. (2015). Making a difference in the teamwork: Linking team prosocial motivation to team processes and effectiveness. Academy of Management Journal, 58(4), 1102-1127.
