'Does Team Research Develop Within Organizational Psychology Over The Years?'

(1)

Organizational Psychology Over The

Years?'

An elaborate review between 1994 and 2006

Master thesis, MscBA, specialization Human Resource Management

University of Groningen, Faculty of Management and Organization

July 5, 2007

REMCO BOSMA

Student number: 1333763

Concordiastraat 15

9741 BA Groningen

Tel: 06-41203592

E-mail: remco_bosma_@hotmail.com

1

st

Supervisor:

Prof. Dr. G.S. van der Vegt

2

nd

Supervisor:

Drs. H. van Polen

Acknowledgement:

(2)

University of Groningen, The Netherlands. 2007 2

ABSTRACT

This review examines research, between 1994 and 2006, relevant to teams which are performing activities in their real environmental setting. The goal was to test past statements about team research. Is team research ‘alive and well’ and is this kind of research developing and getting more sophisticated over the years? These questions were also the research questions of this paper. After analyzing 130 articles from the 9 most prestigious journals within organizational psychology, prove was found for both statements. Over the studied years researchers paid more and more attention to teams. The research questions that were studied also proved to be getting more sophisticated over the years in terms of model type used. Thus, the research questions can both be positively answered.

Key Words: teams, field study, organizational psychology, model type

TABLE OF CONTENT INTRODUCTION 3 THEORY 5 METHOD 9 ` Meta Analysis 9 Selection criteria 9

Categorizations phase and reliability 10

RESULTS 12

Group research trends in organizational psychology 12 Model type trends in Organizational Psychology 14

Moderators and mediators over the years 17

DISCUSSION 19

(3)

INTRODUCTION

Much of the work in organizations is nowadays completed through teamwork: people working together to accomplish outputs beyond the capabilities of people working alone on the task. Success or failure of teams and teamwork can not simply be forecasted by adding up team members talents and available resources, the interaction between team members and other kind of team processes play a crucial role in success or failure. (Marks, Mathieu & Zaccaro, 2001) If we can understand the processes which are used by employees who are working in teams we are able to redesign al kinds of HRM tools in order to let teams work more effective.

But what are teams exactly? There are multiple definitions but in this thesis I will use the one from the work of Hackman in 1987. Cohen and Bailey also used this definition in their work in 1997. This definition is building on the work of Alderfer (1977).

“A team is a collection of individuals who are interdependent in their tasks, who share responsibility for outcomes, who see themselves and who are seen by others as an intact social entity embedded in one or more larger social systems (for example business unit or organization), and who manage their relationships across organizational boundaries.”

Furthermore, I will use the terms ’teams’ and ‘(work) groups’ interchangeable during this paper. This is because popular management literature use most of the time the term ’team’ whereas academic literature tend to use the term ‘group’ (Cohen & Bailey, 1997). So I do not differentiate between the two terms.

The management and academic press increasingly emphasizes the importance of teams for organizational success in the modern economy that is characterized by a need for rapid information exchange and response to customer demands (Cohen & Bailey, 1997) In order to deal with all kind of competitive challenges the use of teams has expanded dramatically. For instance, Gordon (1992) found that 82% of the companies with 100 or more employees used teams one way or another in their organization.

To continue, in the United States productivity and worker flexibility has been increased through the use of work teams. At least 87% of the organizations in the U.S. reported that they use teams one way or another in their business. (Bishop, Scott, Burroughs, 2000).

(4)

turnover, leaner plant structures, and substantial improvements in production cycle time” (Harris, 1992). In sum, nowadays teams are hot.

The research on teams started decades ago. In 1974 Steiner started the debate on team research when he asked himself whatever happened to the group in social psychology. In trying to answer this question Steiner himself performed some meta-analytic research on teams. In 1994 Moreland, Hogg and Hains continued researching this topic and Sanna and Parks followed in 1997. In this thesis I will describe an empirical meta-analytic research on team field studies within organizational psychology between the years 1994 and 2006.

To add something new to the debate about teams I also looked at what kind of moderators en mediators researchers used over these years to answer their research questions, thus, what kind of model type they did use. In doing so I will try to answer the questions whether research on teams is developing over the years or not.

Therefore, the research questions I try to answer are:

(5)

THEORY

Since Steiner in 1974 questioned “whatever happened to the group in social psychologies?” ,a lot happened within group research. Steiner concluded in his article that researchers had lost interest in studying teams but he predicted a revival by the end of the 1970s. However, this revival never came so Steiner switched to a more pessimistic view about the future of group research in his work in 1983.

Follow up research is performed by Moreland, Hogg & Hains in 1994. They reviewed three social psychology journals to determine the nature of group research from 1975 through 1993. They found that group research fell during the late 1970s, remained low with some exceptions in the 1980s, but group research began to rise in the early 1990s. Moreland et all concluded that the increase in the popularity of group research in social psychology can be attributed to the increase of attention to the topic ‘’intergroup relations’’. Interestingly, research after ‘’intragroup topics’’ (ecology of groups, group composition, group structure, conflict within groups and group performance) fell in this period.

Sanna and Parks (1997) continued researching this topic. They tested their assumption whether research on traditional intragroup topics had been shifted to other research areas, organizational psychologists for example. Because organizational psychologists are more outcome oriented and interested in a problem-focuses approach. They concluded that social psychologists tend to research intergroup relations and organizational psychologists intragroup relations. Sanna and Parks finished with two general conclusions: group research is ‘alive and well’, and intragroup research has moved to organizational journals.

This was more or less supported in 2005, by Ilgen, Hollenbeck, Johnson & Jundt

(6)

oriented. But they concluded that since then things have changed, research followed a different path; trying to answer the ‘’why’’ question as well as the ‘’what’’ question.

Thus, according to previous research team research is changing. How did Ilgen et all come to this conclusion in 2005, were they right, and what major developments occurred in the field of team research recently? To answer that question a brief historical overview will be presented about important theory building research that has been conducted. Most recent studies are still one way or another inspired by the findings presented below.

There has been an increased attention on developing theoretical models of team effectiveness, with team processes in a central role the last couple of decades. In these models researchers generally have adopted an input-process-outcome (I-P-O) framework (Marks et al, 2001). McGrath (1984) and Hackman (1987) have done important theory building research which contributed a lot in the team research field. McGraths framework and Hackmans adaptations of this framework are well known. They view processes as mediating mechanisms linking variables such as job design, interdependence, composition and context with outcomes like performance and viability.

This framework has had a powerful influence on recent empirical research, much of which either explicitly or implicitly invokes the I-P-O model (Ilgen et al, 2005).

However, in 1996 Moreland wrote a provocative paper with three specific arguments why this famous I-P-O framework is insufficient for characterizing teams.

First, a lot of the mediating factors that should mediate the inputs into outputs are not processes at all. According to Marks et al (2001) these ‘processes’ are emergent cognitive or affective states. Second, the I-P-O framework implies a single-cycle linear path from inputs through outcomes. Although even the authors of this framework acknowledge that there are potential feedback loops. Finally, the I-P-O framework suggests simple main effect influences proceeding from one category (I, P or O) to the next. Recent research however has reported other interactions as well. For example, I x P interactions and P x P interactions. (Ilgen et al 2005) Based on Moreland (1996).

(7)

explaining variability in team performance and viability. Adding the extra ‘’I’’ at the end of the model explicitly invokes the notion of cyclical causal feedback. Finally, elimination of the hyphen between the letters merely signifies that the causal linkages may not be linear or additive, but rather nonlinear or conditional.” (Ilgen et al, 2005).

On the other hand, Hackman & Morris (1987) stated that research that directly relates measured characteristics of group processes to performance outcomes is scarce. Contrary to Ilgen et all they concluded in 1987 that the question how such mediation takes place remains open and troublesome. Thus, Ilgen et al implicitly say that after 1987 more researchers were interested in deeper and more troublesome relations in group processes, thus in the ‘why’ question.

Now it is interesting to find out whether the unproven statement of Ilgen et al (2005) that, group research is getting more sophisticated the last couple of years, is true? And especially, is this true for organizational psychology mentioned by Moreland in 1994 and Sanna & Parks in 1997?

Does group research evolve over time? Does it become more popular as Sanna and Parks concluded in their paper? Is empirical research on teams, in their real organizational setting, moving to the direction of increased complexity? So, are the hypotheses tested in research changing over time from simple main effect models into moderating, mediating or even moderated-mediated question or models?

Before moving on to the research questions I would like to make clear what I mean with the terms ‘’moderator’’ and ‘’mediator’’.

(8)

other variables, and a mediator variable is something that explains the relationship between the two other variables (Baron & Kenny, 1986).

Research questions

The preceding sections lead to the following research questions:

1. Is team research “Alive and well” within the discipline of organizational psychology? 2. What types of model have been studied within organizational psychology between the years 1994-2006?

- Has this changed over the years? - Is there a clear trend recognizable?

(9)

METHOD

Meta Analysis

To answer the research questions stressed above I and another student reviewed the articles on team research in nine famous organizational journals. These are: Academy of Management Journal (AMJ), Administrative Science Quarterly (ASQ), Journal of Organizational Behavior (JOB), Journal of Applied Psychology (JAP), Journal of Management (JOM), Management Science (MS), Organizational Behavior and Human Decision Processes (OBHDP), Organizational Science (OS) and Strategic Management Journal (SMJ). Sanna and Parks finished their work in 1994, this provided a clear beginning date for this review. We started where they left off. Our research therefore ran from 1994 till 2006. Because the selected journals are well known among researchers we figured that the work in these journals would be representative for much of the research done in the (team research) field over these years. The work of Sanna and Parks also provided excellent guidance for our review; which articles we should include, which criteria to use and so on.

Selection Criteria

Teams which perform their activities in organizational settings (field studies) are the focus in this review. McGrath stated in 1984 that groups should be studied in context so that findings from studies could be more easily generalized to the ‘real world’. The reason behind this is simple; teams performing real tasks in real organizations are influenced by external determinants for effectiveness as well. These external influences are seldom examined in laboratory settings or these are hard to simulate. Thus, “when tasks and norms are held constant (or relatively so), in experimental studies, it is nearly inevitable that the richness and diversity of interpersonal behavior within groups will be reduced substantially.”(Hackman, 1987). So field studies provide a better picture how teams really function.

The attention is restricted to studies where outcomes are measured on the team level. We therefore exclude team studies focusing on individual, business unit or organizational outcomes. This is because the goal of this study is to gain more knowledge on teams, team processes and team outcomes.

(10)

than do other groups, and, some of the phenomena that occur in groups cannot occur at al in dyads.’’ (Moreland et al 1994)

Categorizations phase and reliability

The remaining articles were al analyzed and categorized within the IMOI model from Ilgen et al (2005). The IMOI model is a renewed version of McGrath’s I-P-O model. We further divided the ‘I’ of McGrath’s model into ‘internal’ (group characteristic or task characteristics) and ‘external’ input variables. We spit up the ‘P’ into ‘behavior’ or ‘emergent state’. ‘Performance’, ‘attitudinal’ and ‘behavioral’ were extracted from McGrath’s ‘O’. Furthermore, we checked all articles for ‘control variables’, ‘topic of study’, ‘model type’, ‘location’, ‘industry’ and ‘team type’.

This categorization phase was the main part of our research. All the outcomes were entered into our developed database in SPSS.

Using the selection criteria mentioned above, we reviewed all of the articles found in the nine journal during 1994-2006, and categorized them according to the, by us, adapted framework from McGrath et al.

We categorized the articles in accordance with the framework independently of each other. We then compared our findings. In order to do so we computed Cohen’s Kappa. Cohen’s Kappa is used to assess inter-rater reliability when observing, coding, and so on, qualitative or categorical variables. In doing so takes into account the agreement which can occur by chance. Our overall level of agreement was high, a Cohen’s Kappa of 0,86 for the ‘I, P or O variables’ in the articles and, for my research more important, a Cohen’s Kappa of 0,89 for ‘model type used’ in the articles, indicating good reliability. (See Appendix A)

The discrepancies which arose during the categorization phase were resolved through discussion.

In this paper I differentiate between four types of models: main effect models, moderator models, mediator models and moderated mediated models. We scrutinized for every article which model type is used on team research.

(11)

variable. Moderated mediated models are the most troublesome models; this is because a moderator as well as a mediator is measured which suppose to influence the independent variable. In this sense the moderated mediated model is the most comprehensive model of the four model types.

(12)

University of Groningen, The Netherlands. 2007 12 RESULTS

Group research trends in organizational psychology

Number of team field studies divided over the journals between 1994 and 2006

We first counted the number of the selected articles. Our sample consists at the end of 130 field studies about teams. We then divided the studies over the years and over the selected journals. These results are presented in figure 1 and table 1.

Team field studies

0 5 10 15 20 25 30 2006 2005 2004 2003 2002 2001 2000 1998 1996 1994 Year N u m b e r o f a rt ic le s SMJ OS OBDHP MS JOM JAP JOB ASQ AMJ

Figure 1: Team field studies over the years for the selected journals in figures

2006 2005 2004 2003 2002 2001 2000 1998 1996 1994 Total AMJ 2 7 3 1 6 6 3 2 2 1 33 ASQ 1 0 2 2 1 1 1 0 2 2 12 JOB 3 0 8 4 2 6 1 1 0 0 25 JAP 4 0 2 2 1 1 5 3 2 0 20 JOM 2 2 3 0 1 0 4 0 0 0 12 MS 1 1 3 1 2 0 1 0 0 0 9 OBDHP 0 0 1 1 0 0 1 3 0 0 6 OS 1 2 3 1 1 2 2 0 0 0 12 SMJ 1 0 0 0 0 0 0 0 0 0 1 Total 15 12 25 12 14 16 18 9 6 3 130

(13)

Starting with only a small number of field studies devoted to teams in 1994, the following six years, till 2000, substantially more attention was paid to team research. Then, till 2003, there was a small decrease in the number of studies focusing on teams. In 2004 however, team research reached a peak with 25 studies in that year. This peak can be explained by a special issue that came out in Journal of Organizational Behaviour. This Special issue focussed especially on groups. In the two following years the amount of team studies decreases to the level before the peak. The most popular journal (the journals that paid most attention to team research) turned out to be AMJ, followed by JOB and JAP. The least popular journals were SMJ, OBHDP and MS.

However, to say something about the increase in articles about teams the total number of pages devoted to team field studies is divided by the total number of pages in these journals over these years, in this way my first reseach question can be answered. As a result a proportion index came out. In figure 2 this proportion index is outlined as a figure. Higher proportions indicate higher levels of interest in team field research within the journals in that given year. Looking at figure 2 gives more or less the same picture as figure 1. An increase in interest till 2000, then a small decrease till 2003, a peak in 2004, in 2005 and 2006 the interest level returns to the level just before 2000.

Levels of interest in group field studies from 1994-2006

0 0,01 0,02 0,03 0,04 0,05 0,06 1994 1996 1998 2000 2001 2002 2003 2004 2005 2006 Year In d e x o f in te re s t in g ro u p fi e ld s tu d ie s

(14)

A closer look at the trend line in figure 2 presumes a greater interest in group field research within the discipline of organizational psychology over the researched years. However, to really say something about this line I had to assess whether this trend line is significant or not. In order to do so I looked at the proportions scores and performed time series analysis to test for any effects of years (Y, Y², Y³).Looking at the trend line I assumed a linear effect To find out if the proportion data were normally distributed I performed P-P plots.

Results showed that the data were distributed normally, so a time series analysis was permitted. The time series analysis showed a significant linear trend, F(1,8) = 6,31, p < 0,04. This trend indicates that the interest in team field studies has increased in organizational psychology over the researched years.

Model type trends in Organizational Psychology

Type of model used in team research between 1994 and 2006

Information about the model types studied in articles on groups is shown in Figure 3 and Table 2. The table shows that overall 41,54% of the research studied here used a moderator somehow. This was the most popular model type. Mediation effects were measured in 30% of the cases, followed by 21,54% for main effect models.

The figure clearly stipulates that some models were more frequently used over time than others. For example ‘Moderator models’, researchers started using these models only since 1998. But since then this model became more popular almost every year. The same thing can be said for the ‘Moderated mediated models’, these more sophisticated models were first used in 2000. Since then they were in general more frequently used every year.

(15)

Model types used over the years

0 2 4 6 8 10 1994 1996 1998 2000 2001 2002 2003 2004 2005 2006 Year C o u n t

Main effect model Moderator model Mediator model Moderated-Mediated model

Model not clear

Figure 3: type of model used over the years

2006 2005 2004 2003 2002 2001 2000 1998 1996 1994 Total Main effect model 1 6,67% 2 16,67% 6 24% 4 33,33% 2 14,29% 1 6,25% 7 38,89% 1 11,11% 3 50% 1 33,33% 28 21,54% Moderator model 6 40,00% 4 33,33% 9 36% 3 25% 7 50% 6 37,5% 2 11,11% 4 44,44% 0 0% 0 0% 41 41,54% Mediator model 4 26,67% 2 16,67% 4 16% 3 25% 5 35,71% 7 43,75% 6 33,33% 3 33,33% 3 50% 2 66,67% 39 30% Moderated mediated model 3 20% 4 33,33% 4 16% 2 16,67% 0 0% 2 12,5% 1 5,56% 0 0% 0 0% 0 0% 16 12,3% Not clear 1 6,67% 0 0% 2 8% 0 0% 0 0% 0 0% 2 11,11% 1 11,11% 0 0% 0 0% 6 4,62% Total 15 100% 12 100% 25 100% 12 100% 14 100% 16 100% 18 100% 9 100% 6 100% 3 100% 130 100%

Table 2: Type of models used in the selected articles over the years

Significant trends in’ type of model’ used in team research between 1994 and 2006

(16)

Model type Significance

Main effect model Linear 0,695

Quadratic 0,019* Cubic 0,615

Moderator model Linear 0,001*

Quadratic 0,065 Cubic 0,917

Mediator model Linear 0,848

Moderated-mediated model Linear 0,029*

Not clear Linear 0,486

Quadratic 0,081 Cubic 0,449 Table 3: Type of models used with their level of significance.

* Significant at values < 0.05

I was especially interested in whether the general preference for more elaborated models has increased over time.

A close look at the results in Table 3 teaches us that three results are significant. That means, these are clear trends and cannot be explained due to coincidence or luck.

The main effect model showed a significant quadratic curve. Thus in the early years of this review, as well as in the most recent years, researchers made little use of main effect models. It looks like the glorious days of the use of main effect models are over.

According to Figure 3 and Table 3 the opposite is true for the moderator models. The first moderator models were used in 1998. Since then they became slowly more popular. A linear significance of 0,001 proves this. But it looks like the best have yet to come.

For the mediator models it is not so clear. This model was used since the beginning, became a little more popular in the period 2000-2002 and then remained somewhat stable.

Since no significance was found, we cannot speak about a clear patron.

(17)

What can be said about the studies in which the hypothesis researched were not so easy to label a specific type of model? The amount of studies where the model type was unknown remained stable during the whole period 1994 till 2006. Indicating that there were and that there always will be some studies where it is hard to see what kind of relations exactly is being researched.

In sum, research question 2 can therefore be positively answered.

Moderators and mediators over the years

To go further into detail I also scrutinized what kind of variables were used as moderators and mediators over the years. An overview can be found in Table 4. In this way useful things can be said about what kind of moderators and mediators researchers were interested in over the years and if this has changed. As table 2 already showed no moderators were used in 1994 and 1996, indicating again that moderator models were not used at that time.

A closer look at what kind of moderators were used teaches us that from 1998 till 2004 conflict and disagreement were quite often used as moderators. For example in 2000 Duffy, Shaw and Stark researched the effect of self esteem on group members performance and satisfaction and how relationship conflict influences this relationship. Simons and Peterson studied also in 2000 intragroup trust, task conflict and relationship conflict in top management teams. Furthermore, since 2003 research with empowerment and interdependence used as moderators increased. For instance, Kirkman et al (2004) studied team empowerment in virtual teams with high or low face to face interaction and Colquitt (2004) studied reactions to procedural justice in teams with task interdependence as a moderator.

(18)

It looks like that moderator variables shifted from relationships within teams; conflict and disagreement were used as moderators from 1998 till 2004, to the structure of teams; empowerment and interdependence were the most used moderators since 2003.

The mediators which are used over the years al serve the same goal; researchers try to find answer to the question why some teams perform better than others. They try to explain the differences in performances with these mediator variables.

Year Moderators used Mediators used

1994 - Communication, integration, control

(bureaucratic, personal, group)

1996 - Strategic decision making, conflict (cognitive,

affective), identity (type, strength), group goals, collective efficacy, effective evaluations

1998 Individual self management, time in group, group

size, management style

Conflict, communication, commitment, goal difficulty constructive controversy, team confidence

2000 Relationship conflict, task conflict Cohesiveness, trust, individual mood,

satisfaction (superior, co-workers), commitment (team, organization), commitment to decision

2001 Environmental uncertainty, diversity, group

efficacy, relationship conflict, participation, team design, disagreement, conflict (relationship, task, process)

Cooperative norms, communication (external, internal), shared beliefs, conflict (cognitive, affective), interaction, teamwork agreement, heterogeneity

2002 Task motivation, interpersonal congruence,

relationship conflict, task environment, experience

Information sharing, collaboration, group potency, norms, relational coordination

2003 Power centralization, empowerment, collectivism,

interdependence, relationship conflict, task conflict

Intra-group influence, cohesiveness, reflexivity, relationship (interaction, interdependence), knowledge effort

2004 Empowerment, trust, external context,

cooperation, power distance, task

interdependence, relationship conflict, context

Duration, monitoring, network (range, density), shared vision, team support, knowledge sharing, cohesion, shared mental models

2005 Group potency, individualism/collectivism, task

interdependence, standardized work processes

Conflict (emotional, task), team learning behaviour, team conflict, team cohesion, task interdependence, interpersonal conflict, task conflict

2006 Geographic dispersion, professional status, mental

model similarity, empowerment, autonomy, communication, leadership

(19)

DISCUSSION Conclusion

The results discussed above allow us to draw some general conclusions about team research within organizational psychology. These general conclusions will also answer my research questions. The research questions are:

1. Is team research “Alive and well” within the discipline of organizational psychology? 2. What types of model have been studied within organizational psychology between the years 1994-2006?

- Has this changed over the years? - Is there a clear trend recognizable?

- Does team research in this sense develop?

A goal of this study was to investigate whether there was an increase in team research over the years or not. Figures 1 and 2 clearly stipulate the increase in team research between 1994 and 2006. The significant linear curve indicates that the interest in team research is indeed increasing, that this conclusion is generalizable and coincidence is excluded. This conclusion follows conclusion one from Sanna and Parks in 1997. So there is more and more proof that more attention is paid to team research over the last recent years. In general the attention to team research shows a curve with ups and downs as Steiner already noted in 1972.

In short, I can give a positive answer to research question one.

Another goal of this review was to test statements in past research that team research is indeed changing and developing. In order to answer this second research question I have analyzed what kind of model researchers used over time to test their hypotheses, and whether this has changed over the years. In this paper I differentiated between 4 type of models: main effect models, moderator models, mediator models and moderated-mediated models.

(20)

the future. The same thing can be said about the moderated mediated models. A significant linear trend was found, indicating that the interest in this model type is increasing over the years. The first time this model was used was in 2000, since then it became more popular over the following years. Finally, models which uses mediators. It is hard to say something about the use of this model type over the year. There is no clear significant trend recognizable. The use of this model type remained somewhat stable over the years. In the broader picture, this model type falls out of place.

To conclude: overall the statement that team research is developing over the years can be supported. The fact that research with moderator and moderated mediated models started in 1998 and 2000 indicates that researchers are learning and now try to answer more difficult detailed questions about teams and team performance. In other words, team research indeed develops. The fact that the trend lines are significant and that interest in ‘simple’ main effect models is decreasing, proves this even more. Thus, research question 2 can be positively answered.

However, mediator models form the exception. I expected, like the moderator and moderated mediated models, a significant linear trend indicating more use of this model type over the years. This proved not to be true. The amount of research where it was not clear what relationships are scrutinized, model not clear, remained stable and low over the years. This can be explained in a way that this had no impact on the overall analyses.

After further scrutinizing what variables are used as moderators and mediators in past research I conclude that researchers believe that the variance in team performance can be explained foremost by the following variables; conflict, constructive controversy, communication, commitment, agreement, knowledge sharing, empowerment, interdependence, information sharing and team learning. This is because these variables repeatedly attract the interest of researchers. Another, different, conclusion could be that it is still unknown how these variables influence performance and in what why, so that therefore these variables will come back in research.

Suggestions and contributions

(21)

research questions from what predicts team effectiveness and viability to why some teams perform better than others. They also stated that more attention was paid to mediating processes that explain why certain inputs affect team effectiveness and viability. This statement was not proven with this review.

Thus, Hackman and Morris were right in 1987 with their assumption about mediating variables, that these relationships are hard to scrutinize.

Hackman and Morris also offered three propositions why it is hard to scrutinize mediating processes linking inputs through outputs. First, “the number of factors that can affect group output is so great that managing more than a few factors at a time is nearly impossible. They propose that three ‘’summary variables’’ cause a major portion of variation in group performance: effort, performance strategies and knowledge and skills. Second, each summary variable can be substantially affects, positively and negatively, by what happens in the group interaction process. Third, different summary variables or combinations of them are operative for different types of group tasks. In other words, the variables measured in the mediation process between input and output are heavily determined by the type of the group task performed.” (Hackman and Morris, 1987)

This might explain why the mediator models showed not the same (significant) trend as the moderator and moderated mediated models. First there should be more knowledge available about the summary variables, mentioned by Hackman and Morris, and their relationship with team performance and viability.

Other research is needed to scrutinize what is already known about these variables and what is discovered about relationships with performance and viability.

It sounds logical that when these questions are answered the use of mediated models will dramatically increase.

(22)

moderators and mediators. If this is not the case then I expect that researchers will no longer use these variables so often as moderators and mediators anymore.

The most important contribution of this research to the knowledge on group research is that for the first time since 1994 research trends on group field studies have been empirically explored. Moreland et all finished their review in 1994 and that is why I chose 1994 as starting point. This thesis has proved that the interest in group field research has increased since 1994. Another contribution should be that it delivered some solid indications and trends, which should give other researchers in this field some inspiration to do follow up research or continue with this one.

A drawback of this review is that we can only draw conclusions on field studies since our sample only consist of these kinds of studies. I can only hypothesize that the same trend can be found for other methodologies and other disciplines then organizational psychology. In order to prove this, new research should include these other methodologies and disciplines. To make some genuine comparisons with this study, the same methods should be used.

Another drawback is the relative small sample of articles studied in this review (N=130). This can be explained by the fact that this study only focussed on field studies. Thus, the follow up research I recommended would not have this drawback.

(23)

REFERENCES

• Baron, R. M., Kenny, D. A. 1986. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51: 1173-1182.

• Bishop, J. W., Scott, K. D., Burroughs, S. M. 2000. Support, commitment and employee outcomes in a team environment. Journal of Management, 26: 1113-1132.

• Cohen, S. G., Bailey, D. E. 1997. What makes teams work: group effectiveness research from the shop floor to the executive suit. Journal of Management, 23:239-290.

• Colquitt, J. A. 2004. Does the justice of the one interact with the justice of the many? Reactions to procedural justice in teams. Journal of Applied Psychology. 89:633-646

• Cummings, J. N. 2004. Work groups structural diversity and knowledge sharing in a global organization. Management Science. 50:352-364.

• Duffy, K. M., Shaw, J. D., Stark, E. M. 2000. Performance and satisfaction in conflicted interdependent groups: when and how does self esteem make a difference? Academy of Management Journal, 43:772-782

• Gordon, J. 1992. Work teams: how far have they come? Training, (October): 59-65.

• Hackman, R. J., Morris, C. G. 1987. Group tasks, group interaction process, and group performance effectiveness: A review and proposed integration. Handbook of organizational behaviour: 45-99. Englewood Cliffs, NJ: Prentice Hall.

• Harris, T. E. 1992. Toward effective employee involvement: An analysis of parallel and self -managing teams. Journal of Allied Business Research, 9 (1): 25–33.

• Ilgen, R. D., Hollenbeck, J. R., Johnson, M., Jundt, D. 2005. Teams in Organizations: from input-process-output models to IMOI models. Annual Review of Psychology, 56:517-543.

• Kirkman, B. L., Rose, B., Tesluk, P. E., Gibson, C. B. 2004. The impact of empowerment on virtual team performance. Academy of Management Journal. 47:175-192.

• Marks, M. A., Mathieu, J. E., Zaccaro, S. J. 2001. A temporally based framework and taxonomy of team processes. Academy of Management Review, 26: 356-376.

(24)

• Moreland, R. L., Hogg, M. A., Hains, S. C. 1994. Back to the future: Social psychological research on groups. Journal of Experimental Social Psychology, 30: 527-555.

• Sanna, L. J., Parks, C. D. 1997. Group research trends in social and organizational psychology: whatever happened to intragroup research? American Psychological Science, 8: 261-267

• Simons, T. L., Peterson, R. S. 2000. Task conflict and relationship conflict in top management teams, the pivotal role of intragroup trust. Journal of Applied Psychology. 85:102-111

• Steiner, I. D. 1972. Group process and productivity. New York: Academic Press.

(25)

Appendix A: Cohen’s Kappa

Model type 2003 Remco Bosma 1 2 3 4 5 Row total 1 4 4 2 2 2 3 1 3 4 4 2 2 5 0 Column total 4 3 3 2 0 12

Total number of agreements: 4 + 2 + 3 + 2 = 11

Frequencies of agreement by chance: ef = (Row Total * Column Total) / Overall total

Cohen’s Kappa: ∑a - ∑ef 11 – 3,167 K= --- = --- = 0,887 N - ∑ef 12 – 3,167 Model type 2005 Remco Bosma 1 2 3 4 5 Row total 1 2 2 2 3 1 4 3 2 2 4 4 4 5 0 Column total 2 3 3 4 0 12

Total number of agreements: 2 + 3 + 2 + 4 = 11

Frequencies of agreement by chance: ef = (Row Total * Column Total) / Overall total Cohen’s Kappa:

∑a - ∑ef 11 – 3,167

K= --- = --- = 0,887

N - ∑ef 12 – 3,167

Average Cohen’s Kappa over 2003 and 2005 (type of model): 0,887

(We assume this is representative for the other years as well.)