Master Thesis
Nanxi Chen s2271966
Technology & Operations Management
nancychen9077@gmail.com
Communication patterns, task interdependence and group
- 1 - CONTENT 1. INTRODUCTION ... - 3 - 2. THEORETICAL BACKGROUND ... - 5 - 2.1. GROUP COHESION ... - 6 - 2.2. COMMUNICATION PATTERNS ... - 7 -
2.3. TASK INTERDEPENDENCE IN GROUP DECISION MAKING ... - 9 -
2.4. CONCEPTUAL MODEL AND HYPOTHESES ...- 11 -
3. METHODOLOGY ... - 15 -
3.1. OVERVIEW ... - 15 -
3.2. CONTROLLED EXPERIMENT ... - 15 -
3.2.1. Experimental Design ... - 15 -
3.2.2. Experiment Task ... - 15 -
3.2.3. Treatment Factors and Level ... - 16 -
3.2.4. Manipulation... - 16 -
3.2.5. Measurement ... - 17 -
3.2.6. Experimental Settings ... - 21 -
4. DATA ANALYSIS AND RESULT ... - 22 -
4.1. RELIABILITY CHECK ... - 22 -
4.1.1. Perceived Task Interdependence ... - 22 -
4.1.2. Group Cohesion ... - 22 -
4.2. MANIPULATION CHECK ... - 24 -
4.2.1. Communication Patterns ... - 24 -
4.2.2. Perceived task interdependence ... - 24 -
4.3. HYPOTHESIS TEST ... - 26 -
4.3.1. Task interdependence and group cohesion ... - 26 -
4.3.2. Communication Patterns and group cohesion... - 28 -
4.3.3. Communication Patterns and Perceived task interdependence (PTI) - 32 - 4.3.4. Interaction Effects ... - 34 -
5. DISCUSSION ... - 39 -
5.1. DISCUSSION OF THE RESULTS ... - 39 -
5.2. LIMITATION ... - 40 -
5.3. CONTRIBUTION TO THEORY AND MANAGERIAL RELEVANCE ... - 41 -
5.4. DISCUSSION OF FUTURE DIRECTIONS ... - 41 -
6. CONCLUSION ... - 43 -
- 2 -
Abstract
Nowadays, work groups have become more and more important within organizations. Existing literature suggests that group cohesion, communication patterns and task interdependence are all vital factors closely related to group process and outcome. However, there are only limited works with regard to the interrelationships among those factors, especially for the conflicting role of task interdependence, whether it serves as a mediator or a moderator between the relationship of communication patterns and group cohesion. In this study, a controlled experiment is conducted to explore the interrelationship among the three factors. The results indicate that task interdependence can positively affect the degree of group cohesion. Communication patterns have no significant impact on both task interdependence and group cohesion. The findings are discussed in terms of theoretical and managerial relevance and possible future studies.
Key words: Communication Patterns, Perceived Task Interdependence,
- 3 -
1. Introduction
Nowadays, work groups have become more and more important within organizations. As a result, research concerning work group cohesion receives a lot of attention, because it can help to strengthen organizational group decision making processes and outcomes (Summers et al. 1988). Therefore, to better understand group decision making processes and outcomes, it is necessary to investigate issues with regard to group cohesive within a work group. In this study, the groups all refer to work groups.
According to the themes and characteristics model provided by Campion & Medsker (1993), task interdependence and communication pattern are vital factors directly related to work group effectiveness. More specifically, work group effectiveness can be influenced by communication patterns (Goodman et al. 1986). That is to say, group effectiveness calls for the match of communication patterns and the nature of group tasks. And they found out that the higher the requirement of group effectiveness, the higher the level the communication patterns should match the group task nature. Meanwhile, task interdependence mediates the interactive effects of diversity and cohesion on group efficacy (Rogelberg & Rumery 1996).
In general, group cohesion refers to the sense of belongingness that an individual feels for a group (Furumo & Pearson 2006). Communication patterns refer to face-to-face or computer-mediated modes of exchanging information. Task interdependence refers to the degree to which group members rely on each other to perform their tasks effectively given the design of their jobs or the interconnection between the tasks of group members(Saavedra et al. 1993).
- 4 -
communication patterns. The relationship becomes rather complex when task interdependence is also considered. The confusing role of task interdependence, whether it serves as a moderator or mediator with respect to the relationship between communication patterns and group cohesion, still remains unclear in previous literature (Straus & McGrath 1994; Rhoads 2010). For instance, it is closely related to both group task norms and task-oriented group communications, which both have great impact on group cohesion and communication patterns. Hence, it might also have complex impact on group cohesion and communication patterns. Therefore, to better understand the work group decision making process and outcomes, it is critical to have a better understanding of the interrelationship among the three factors, group cohesion, communication pattern, and task interdependence.
The research question is:
How are the communication patterns, task interdependence and group
cohesion related to each other in work groups?
The objectives are:
To explore the relationship between communication patterns, task interdependence and group cohesion in work groups with controlled experiments;
To discuss the management and theoretical implications based on the comparison of empirical results in work group development.
- 5 -
experiment task. Further study should concern about some other factors, for instance, the communication frequency. Different communication medium might not only lead to different communication patterns but also different communication frequencies.
The paper is structured as follows. In the next section, the theoretical background, conceptual model and hypothesis are provided. Then, in section 3, the methodology is given, which is about setting of the experiment, manipulation of variables and the analysis of the expected results. In section 4, the data analysis was conducted including the manipulation check. Afterwards, a discussion part is organized to analyze the findings and address the limitation and the implications of the study. Finally, the study's conclusion is made.
2. Theoretical Background
Work groups are made up of individuals functioning as social entities, and individuals jointly deal with interdependent tasks as members of such groups (Guzzo & Dickson 1996). By far, considerable studies have been conducted to better understand such organizational work groups in psychological and behavioural perspectives. As reviewed in Goodman et al. (1986), different variables are investigated to better understand coordination, group effectiveness and group performance. Hence, the main criteria to decide whether it makes sense to study a variable concerning work groups, depends on its potential impact on group coordination, group effectiveness, group efficacy and group performance, which are all closely related to group decision-making process and outcomes.
- 6 -
Based on previous work, Goodman et al. (1986) concluded that group effectiveness can be influenced by communication patterns. For computer-mediated and face-to-face groups, the communication patterns are different. Both are widely used for quite a long time.
Task interdependence functions as a major determinant of group coordination (Van de Ven et al. 1976) and group tasks (Gladstein 1984).As a result, it can moderate the relation between group decision making process and group effectiveness (Gladstein 1984). Moreover, it may also mediate the interactive effects of group diversity and group cohesion on group efficacy (Rogelberg & Rumery 1996).
In general, group cohesion refers to the sense of belongingness that an individual feels for a group (Furumo & Pearson 2006). Communication patterns refer to the mode group members interact with each other. Task interdependence refers to the degree to which group members rely on each other to perform their tasks effectively given the design of their jobs or the interconnection between the tasks of group members (Saavedra et al. 1993).
As indicated above, all of the above variables are relevant to group work processes and outcome. Additionally, due to the multidimensionality of group performance, it is complex and difficult to evaluate work groups (Chang & Bordia 2001). To make it easier, based on evidence from meta-analysis, group cohesion may facilitate performance and has consistent effects in a wide variety of settings and tasks (Gully et al.1995). By investigating the potential influence of communication patterns and task interdependence on group cohesion, we can better evaluate group performance with regard to communication patterns and task interdependence.
Hence, previous works concerning group cohesion, communication patterns and task interdependence are firstly reviewed. Afterwards, based on existing findings of the influences, a conceptual model is built and hypotheses are provided.
2.1. Group cohesion
- 7 -
feelings of belongingness or attraction to the group” (Delbecq 1974). Obviously, this refers to a psychological force that binds people together (Keyton & Springston 1990). Moreover, it is an important element of group dynamics (Evans & Jarvis 1980). Scholars have been dedicated to investigate different aspects of group cohesion, and their findings about personal change groups are consistent. Group cohesion and group member outcome are proved to be positively related (Stokes 1983).
In Evans & Dion (1991), the researchers investigated the relationship with a meta-analysis. With the help of 27 basic studies, the findings suggest that group cohesion is positively related to group performance. Then, in a later work by Hogg (1992)
,
the scholars state that cohesive group members exhibit more positive, personal, and favourable communication interactions. Based on a review of existing literature, Yoo & Alavi (2001) conclude that cohesive groups will have sociable, warm, and personal interactions between members, thus enhancing the social presence of communication interactions. More specifically, in their work, they prove that communication medium and group cohesion are related.However, there are also considerable conflicting arguments with regard to groupthink. Some authors indicated that too much group cohesiveness may possibly result in poor group performance (Janis 1971). Fortunately, group cohesion is only a necessary but not a sufficient factor that leads to groupthink. (Mullen et al. 1994; Bendoly et al. 2009). More specifically, in Bernthal & Insko (1993), the findings proved that groupthink is least likely to appear when task-oriented group cohesion exceed social-emotional group cohesion. To decide if it is the double edged sword,
Langfred (1998) conducted a survey and indicates that group task norm moderates the relationship between performance and cohesion.
Therefore，due to its conflicting impacts on group performance, it is critical to understand the group cohesion before going further.
2.2. Communication patterns
- 8 -
experienced great change. Face-to-face communication used to be the only pattern for group decision making. Then, due to advanced computing technologies, computer-mediated communication emerged. Now, both patterns are jointly facilitating group decision making in organizations.
In fact-to-face pattern, group members are collocated and exchange information directly. Alternatively, in computer-mediated pattern, group members are separated in geographically different places and are exclusively allowed to communicate using computer techniques, for instance email and software like MSN. Virtual teams are defined as ‘groups of geographically, organizationally organization and/or time dispersed workers brought together with information and telecommunication technologies to accomplish one or more tasks’ ( Powell et al. 2004). The major communication pattern for such groups is the computer-mediated pattern. Additionally, information transmitted with different communication patterns, refers to both task and interpersonal messages (Rhoads 2010).
When it comes to the impact of communication patterns on group performance, it is suggested that there is a trade-off between ease of interactions and effectiveness of collaboration, considering the effect of proximity (Hawkey et al. 2005). More specifically, in group decision making, the ease of interactions influences the feasibility of a decision, while the effectiveness of collaboration affects the flexibility of a decision. Therefore, with different influences on the two factors, different communication patterns may lead to different group performance.
- 9 -
In a later work by Kennedy et al. (2010), the scholars take three communication patterns into account, with an additional mix media communication patterns. The findings present that face-to-face group might keep better group outcomes over time. By conducting a first session face to face could lead to better outcomes in future computer-mediated interactions. However, there’s no significant difference using mix compared to pure computer-mediated and pure face-to-face patterns. It seems in this work, the priority of face-to-face group is addressed.
According to Campion & Medsker (1993), communication patterns and task interdependence are both vital factors directly related to work group effectiveness and group performance.
2.3. Task Interdependence in group decision making
Task interdependence refers to the ‘‘degree to which an individual’s task performance depends on the efforts or skills of others.’’(Wageman & Baker 1997) It is determined by the type of a task and the technology to complete this task
(Wageman 1995). Thus, Vijfeijken & Kleingeld (2002) define it as the degree to which group members have to exchange information and/or means for the completion of the group task in a later work. And it is believed that the degree of task interdependence increases as the difficulty to perform the task increases ( Van Der Vegt & Vliert 2002).
It may affect the group decision making process in organizations in three different aspects. Firstly, it may pose significant challenges for successful implementation of decisions. Secondly, it may affect outcome interdependence by shaping the reward system. Finally, it is critical in shaping organizational mechanisms.(Rajeev & Yetton 2003)
- 10 -
group members perceive task interdependence, they are motivated to interactive coordinate with each other.
According to Ramamoorthy & Flood (2004), when the tasks are highly interrelated, or as long as individuals perceive that tasks are highly interrelated, task interdependence exists. Furthermore, they argue that perceived task interdependence moderates the relationship between loyalty and pro-social behaviours. Perceived task interdependence could serve as a possible measure technique to evaluate task interdependence (De Snoo & Van Wezel 2011).
- 11 -
2.4. Conceptual model and hypotheses
Although the influence of those three factors above on group process and outcome received a lot of concern in research, there are only limited works with regard to the interrelationships among those factors.
The following conceptual model and hypotheses are based on the theory concerning testing the mediation and moderation effect provided by Baron & Kenny (1986). According to them, moderating effects exist if the effects illustrated by Path 1 and Path 3 in Figure 1(a) exist, and it doesn’t matter if the effect of Path 2 is significant. Mediating effects exist when the effects illustrated by Path4, Path 5 and Path 6 in Figure 1 (b) exists. The moderating effect is significant when the effect of Path 3 is significant. But the mediating effect is significant when the previous significant effect of Path 4 no longer exist, if the effects of Path 5 and Path 6 are controlled.
.
Figure 1 (a) Moderating Effect
Figure 1 (b) Mediating Effect
- 12 -
predictor here, and group cohesion is the outcome variable (Baron & Kenny 1986). To determine the relationship between communication patterns and group cohesion, the role of task interdependence should be firstly be evaluated.
Task interdependence is characterized as one of the main causes of the inconsistent finding relates to group cohesion and group performance (Gully et al. 1995). More explicitly, a strong cohesive-performance relationship exists for interdependent task. Hence, it functions as a moderator of the relationship of group cohesion and performance.
However, its impact on group cohesion alone remains unclear. Face to face communication group is proved to be more productive than computer mediated group, in high task interdependence conditions (Straus & McGrath 1994). This finding suggests that task interdependence might serve as moderator of the relationship between communication patterns and group outcomes. Alternatively, according to
Rhoads (2010),task-oriented communication is more important in virtual groups, and this is due to the limitation on communication patterns. It is possible that due to the limitation of communication patterns, group members perceive different level of task interdependence. Hence, it can be deduced that task interdependence can be influenced by communication patterns. In this case, task interdependence might function as a mediator.
As indicated above, the role of task interdependence, whether it serves as a moderator or mediator with respect to the relationship between communication patterns and group cohesion, still remains unclear in previous literature. Hence, to better understand the relationship among the three variables, it is critical to identify the function of task interdependence in this relationship.
Therefore, the conceptual model and hypotheses are then generated to explore the relationship among communication patterns, task interdependence and group cohesion.
- 13 -
investigate a more promising relationship, hypotheses with regards to each possibility should be generated and tested separately. The main distinction is the confusing role of task interdependence.
Two hypotheses should be tested firstly, regardless of the confusing impact of task interdependence.
Figure 2. Conceptual Model
The degree of perceived task interdependence positively affects the level of group cohesion.( This is formulated based on the effects depicted by Path2 and Path6 in figure 1 (a) and Figure1 (b) respectively.)
Hypothesis 1 The higher the degree of perceived task interdependence, the higher the level of group cohesion.
Communication patterns have an influence on group cohesion. (This is formulated based on the effects depicted by Path1 and Path4 in Figure 1 (a) and Figure1 (b) respectively.)
Hypothesis 2 Face-to-face communication results in higher group cohesion, compared to computer-mediated communication.
To determine if task interdependence serves as mediator, the following hypotheses should be tested.
Communication patterns have an influence on perceived task interdependence. (This H 1
H 2 H 4 H 3
- 14 -
is formulated based on the effects depicted by Path5 in Figure1 (b).)
Hypothesis 3 Face-to-face communication results in higher degree of perceived task interdependence, compared to computer mediated communication.
When perceived task interdependence is at a fixed level, communication patterns have less effect on group cohesion. (This is formulated based on the effects depicted by Path4 when the effects of Path5 and Path 6 in Figure1 (b) are controlled.)
Hypothesis 4 When the levels of perceived task interdependence are the same, group cohesion in a face-to-face group is similar to that on a computer-mediated group.
To determine if task interdependence serves as moderator, the following hypothesis should be tested.
The communication patterns and perceived task interdependence have combined influence on group cohesion. (This is formulated based on the effects depicted by Path3 in Figure 1 (a).)
Hypothesis 5 Perceived task interdependence moderates the impact of communication patterns on group cohesion.
- 15 -
3. Methodology
3.1. Overview
To investigate the relationship between communication patterns, task interdependence, and group cohesion, controlled experiment is the most appropriate method. It is generally found that the settings and conditions can be better controlled in laboratory than in a field study (Bendoly et al. 2006). The following part is for experiment design.
3.2. Controlled Experiment
3.2.1. Experimental DesignFour practical scenarios are simulated with a 2*2 experimental design, illustrated in
Figure 3. (As mentioned in the literature review, the task interdependence studied
refers to perceived task interdependence)
Figure 3 experimental design
3.2.2. Experiment Task
The experiment task is a scheduling task. There are three main rationales for choosing this task.
Firstly, it is a complex task, which calls for group decision making. It has been proved that interaction and coordination is a critical factor that has a decisive impact on the outcome of a scheduling task.
- 16 -
scheduling activities. More specifically, many scholars have identified the gap caused by ignorance of human factors and conduct various controlled experiments to better understand this type of group decision making activities. As a result, the proposed experiment can borrow some ideas from the previous ones.
More importantly, scheduling has a direct and significant impact on operational performance in practice. As a result, by investigating the proposed relationship in this task, there will be more practical suggestions on improving scheduling performance.
3.2.3. Treatment Factors and Level
Treatment factors are the factors that will be manipulated explicitly in the experiment. In this experiment the treatment factors are communication patterns and the degree of task interdependence.
As depicted in Figure 3 above, the levels of communication patterns are ‘computer-mediated’ and ‘face-to-face’; and the levels for the task interdependence are ‘low’ and ‘high’. The levels of communication patterns were measured between subjects where one group was exposed to computer- mediated communication patterns and the other was exposed to face-to-face communication. The task interdependence levels were measured within the subjects so that each group participates in an experiment composed of two parts, one with low task interdependence and one with high task interdependence. The reason that not all factors were measured within subjects is to avoid learning effect.
3.2.4. Manipulation
Manipulation Communication patterns
- 17 -
Manipulation Perceived Task Interdependence
The manipulation is based on the following theory. The degree of task interdependence may increase as the difficulty to perform the task increases (Van Der Vegt & Vliert 2002). It is determined by the type of a task and the technology to complete this task (Wageman 1995). In addition, it can be strengthened by training and shifting tasks (Rico et al. 2009). Therefore, it is possible to manipulate task interdependence.
In this experiment, the manipulation of perceived task interdependence was realized with the help of different tasks. More specifically, in this study two different types of task concerning scheduling were considered. One is scheduling task and the other is rescheduling task.
According to De Snoo & Van Wezel (2011), coordination exist in both scheduling and rescheduling tasks.However, in scheduling, there is sufficient time for schedulers to make decisions; while in rescheduling, schedulers always experience great time pressure. Hence, scheduling and rescheduling are two different types of tasks. In addition, in scheduling task the schedulers are jointly responsible for the performance of the whole factory, while in rescheduling tasks, schedulers are judged only by the performance of their own department. More specifically, due to the goal shifts, the task interdependence can be changed in different tasks. That is to say, the participants can perceive higher task interdependence in scheduling task than in rescheduling task.
Therefore, in this experiment, scheduling task is conducted in the levels with high task interdependence; rescheduling task is conducted in the levels with low task interdependence.
3.2.5. Measurement
Task interdependence
- 18 -
is finished.
Questionnaire: (Answers were provided on a Likert 7-point scale ranging from “not
dependent” to “fully dependent”).
“How dependent were you on your partner in order to carry out your task adequately”
Then the same measurement of perception of the level of dependence was applied.
“How dependent was your partner on you in order to carry out his or her task adequately.”
Both questions were asked after each task.
Group cohesion
As indicated in the last section, the scholars reach consensus on the importance of group cohesion on group performance. However, due to the complex nature of group cohesion, the measurement of group cohesion remains difficult. More specifically, variables such as inter member attraction, instrumental value, risk taking, group diversity and even group size, have great impact on group cohesion.(Stokes 1983; Carron & Spink 1995; Harrison et al. 1998)
In Siebold (1985) , the author reviewed former work concerning the measurement of group cohesion. He also stated that the measurement of cohesion was due to the different definition of cohesion, the inconsistent professional situation to conduct cohesion research, and also the various technology and methods available. As a result, in this study, to identify an appropriate measurement the three aspects mentioned above were considered carefully. Then in a later work ( Carron & Brawley 2000), again the authors emphasized the importance to study the cohesion of different types of groups in different social contexts.
Accordingly, the following issues were also considered. Firstly, in this study, the group cohesion refers to work cohesion, which is closely related to organizational issues (Summers et al. 1988). Secondly, the Group Environment Questionnaire
- 19 -
cohesion have not been checked. In addition, in this study small group cohesion should be considered, since the experiment task involved small groups. As indicated in Carron & Spink (1995), with changes in group size, the social psychology of individual group members changes a lot. Furthermore, the group cohesion to measure in this study should also take the perceived group cohesion into account.
Bollen & Hoyle (1990) defined perceived group cohesion as “an individual’s sense of belonging to a particular group and his or her feelings of morale associated with membership in the group“. Based on this definition, they also put forward an appropriated technique to measure perceived group cohesion, the Perceived Cohesion Scale (PCS). It is brief, suitable for a broad range of groups and in line with the definition above. Moreover, Chin et al. (1999) adopt this approach and tested it in a small group setting, and its validity and reliability has been proved to be suitable for small groups. Hence, this PCS is also appropriate in this study.
Table 1 Perceived Cohesion Scale
Perceived Cohesion Scale
Sense of Belonging
Q1 I feel a sense of belonging to this group. Q2 I feel that I’m a member of this group. Q3 I see myself as part of this group.
Feeling of Morale
Q4 I am content to part of this group. Q5 I am happy to be as part of the group. Q6 This group is one of the best anyway.
- 20 -
Bordia 2001). In addition, task interdependence is regarded as an important experimental variable. Therefore, in this study a part of GEQ to measure task cohesion should also be used, but it is not necessary to test the social cohesion here.
Table 2 GEQ-Task Cohesion
GEQ-Task Cohesion
Individual Attraction to the Group Task
Q7 I’m not happy with the amount of workload I paid to accomplish the task. Q8 I’m unhappy with my group’s level of commitment to the task.
Q9 This group does not give me enough time to improve my performance. Q10 I do not like the style of work in this group.
Group Integration Task
Q11 Our group is united in trying to achieve its goal.
Q12 We all take responsibility for any poor group performance.
Q13Our group members have conflicting aspiration for group performance.
Q14 If members of our group have problems in fulfilling task, everyone wants to help them so we can work together again.
Q15 Members of our group do not communicate freely about the correct method to accomplish the task.
The feasibility to measure group cohesion in such short controlled experiment was also considered. It is enough to produce some cohesion, and the more time people spend together, the stronger their cohesion becomes (Levine & Moreland 1990). This
indicates that when people are engaged in a group, group cohesion can be generated. Therefore, even in a time consuming experiment situation, group cohesion can be produced.
- 21 -
inter set discriminating power, 9 or more is better. In this experiment, the inter set discriminating power is quite important, by considering those factors above, 9 Likert-type scale was more suitable.
3.2.6. Experimental Settings
In this experiment, two participants worked as a group. They were schedulers of a furniture factory, which has a flexible job shop with four manufacturing departments. Each scheduler was responsible for two departments, and each department consists of two identical machines. For different products, the production route and processing time are different. The schedulers had to determine for which operation which machine is used and at what time it is performed. During the scheduling phase, each scheduler created valid and efficient schedules for his own department, and the scheduler’s goal was to minimize the cost generated by due date violation and unscheduled operation of the whole factory. However, in the rescheduling phase, each scheduler had to adjust the schedule due to unexpected events, for instance, rush order or maintenance, but the scheduler’s goal was to minimize the cost in his own two departments. The Gant-chart was used in conducting the experiment.
Due to the within subject and between subject design, 20 students participated in the experiment. For each subject group, 10 participants were involved.
- 22 -
4. Data Analysis and Result
4.1. Reliability Check
4.1.1. Perceived Task Interdependence
Two questions were included in the questionnaire to test the PTI, and each provides a score ranging from 1 to 7. There were two questions related to the perceived task interdependence. Since both questions were positively related to PTI, each rate was considered an initial score.
To begin with, the scales of Trust data were checked firstly with the Cronbach’s alpha. The related Cronbach’s alpha value is 0.947 and it is higher than 0.7, which indicates a high level of internal consistency to measure PTI. Therefore, a weighted score for PTI was calculated by the following formula.
PTI=
For instance, if one participant rates the first question 6 and the second question 7, the PTI should be 92.86. (
). All the original data gathered for PTI during the experiment is calculated in this manner.
4.1.2. Group Cohesion
As illustrated in the methodology part, two main criteria, representatively the perceived group cohesion (PGC) and group environment questionnaire (GEQ) were applied to measure the group cohesion. For each criterion, multiple questions were involved to measure the group cohesion, 6 for PGC and 9 for GEQ respectively. Moreover, since the two criteria refer to different aspects of group cohesion, a third criterion is added to measure the total group cohesion (TGC). Thus all the 15 questions are considered with regard to TGC.
For the questions asked in the questionnaire, some of them are positively related to group cohesion, while others are negatively related to group cohesion. Since the data gathered is nominal, for each question a score from 1 to 9 is used to translate the score to ordinal data. The rule is the higher the group cohesion, the higher the score.
- 23 -
different criterion. The results in Table 3 suggest that there exists a high level of internal consistency to measure all of the three criteria.
Table 3 Crobach’s alpha results
PGC Cronbach's Alpha and Item-Total Statistics TGC Cronbach's Alpha and Item-Total Statistics
Cronbach's Alpha Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted Cronbach's Alpha Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted Q1 0.960 0.880 0.954 Q1 PG C 0.943 0.775 0.938 Q2 0.918 0.948 Q2 0.758 0.938 Q3 0.934 0.946 Q3 0.803 0.937 Q4 0.927 0.947 Q4 0.868 0.936 Q5 0.890 0.951 Q5 0.912 0.934 Q6 0.781 0.970 Q6 0.826 0.935
GEQ Cronbach's Alpha and Item-Total Statistics Q7
GE Q 0.452 0.945 Cronbach's Alpha Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted Q8 0.658 0.940 Q7 0.887 0.446 0.889 Q9 0.760 0.937 Q8 0.707 0.872 Q10 0.819 0.936 Q9 0.784 0.863 Q11 0.750 0.937 Q10 0.835 0.856 Q12 0.788 0.937 Q11 0.710 0.869 Q13 0.585 0.942 Q12 0.742 0.867 Q14 0.739 0.938 Q13 0.559 0.882 Q15 0.454 0.950 Q14 0.616 0.877 Q15 0.518 0.893
Moreover, when the ‘Cronbach's Alpha if Item Deleted’ is considered, it is obvious that it
- 24 -
a higher level of Cronbach’s Alpha can be gained. Admittedly, the differences after removing the two questions are quite small, it is still better to remove them when a more accurate measurement of group cohesion is needed. Question 6 cannot be removed to measure PGC because it is necessary to take that into account to measure TGC.
Hence, the weighted scores for each criterion were calculated with the following formulas. PGC =
GEQ=
TGC=
For instance, if the scores for Q1 to Q6 are 8, 7, 5, 6, 6, 9, then the PGC value should be
75.93. ( ). All the original data gathered for PGC, GEQ and TGC during the experiment is calculated in the above manners.
4.2. Manipulation Check
4.2.1. Communication PatternsThe communication patterns in the experiment are restricted by the distance between group members in the group. In the face-to-face communication pattern, the schedulers were located right next to their partner so that they can talk freely. In the computer-mediated communication pattern, the schedulers were located separately so that they can only reach their partner through a chat window. Hence, the manipulation of the treatment factor -communication patterns and its treatment level is guaranteed.
4.2.2. Perceived task interdependence
- 25 -
Table 4 Descriptive Statistics for PTI
Descriptive Statistics
N Mean
Std.
Deviation Minimum Maximum PTI Scheduling 20.00 84.29 11.95 57.14 100.00 PTI Rescheduling 20.00 70.36 24.66 28.57 100.00
According to the descriptive statistics data, the average perceived task interdependence is higher in scheduling task than that in rescheduling task. Moreover, the deviation is higher in rescheduling task than that in scheduling task. Then, statistic tests are used.
Table 5 Normality Test for PTI
Tests of Normality Task Shapiro-Wilk Normality Statistic df Sig. PTI Scheduling .872 20 .013 not Rescheduling .867 20 .010 not
The normality test is firstly applied and it turns out the PTI of scheduling task and rescheduling task are neither normally distributed. Moreover, for the scheduling and rescheduling task, the experiment was done within subjects, which suggests that the two samples are related. Hence, Wilcoxon signed rank test was used to investigate the relationship and the results are depicted below.
- 26 -
Table 6 Wilcoxon Signed Ranks Test for PTI
Wilcoxon Signed Ranks Test
N Mean Rank Sum of Ranks Z Asymp. Sig. (2-tailed) PTI Rescheduling – PTI Scheduling Negative Ranks 10a 8.05 80.50 -2.458 .014 Positive Ranks 3b 3.50 10.50 Ties 7c Total 20
a. PTI Rescheduling < PTI Scheduling
b. PTI Rescheduling > PTI Scheduling
c. PTI Rescheduling = PTI Scheduling
4.3. Hypothesis Test
4.3.1. Task interdependence and group cohesion
Table7 Descriptive Statistics PGC GEQ and TGC
Descriptive Statistics
N Mean Std. Deviation Minimum Maximum PGC Scheduling 20 83.9815 10.51478 61.11 100.00 GEQ Scheduling 20 79.7619 15.12765 46.03 100.00 TGC Scheduling 20 81.7094 12.27165 52.99 99.15 PGC Rescheduling 20 71.6667 18.16525 14.81 90.74 GEQ Rescheduling 20 71.6667 16.62604 30.16 88.89 TGC Rescheduling 20 71.6667 16.78597 23.08 88.89
According to the descriptive statistics data, under each criterion the average group cohesion is higher in scheduling task than that in rescheduling task. The deviation is higher in rescheduling task than that in scheduling task. Moreover, the average scores of group cohesion in rescheduling are similar in all three criterions.
- 27 -
Table 8 Normality Test for Task--Group Cohesion
Tests of Normality Task Shapiro-Wilk Normality Statistic df Sig. PGC Scheduling .967 20 .688 yes Rescheduling .729 20 .000 not GEQ Scheduling .930 20 .158 yes Rescheduling .904 20 .048 not TGC Scheduling .954 20 .434 yes Rescheduling .799 20 .001 not
According to the normality test result, Wilcoxon signed rank test is more appropriate to explore the relationship here for each criterion since the samples are related and the findings are depicted below. In this case, single tailed tests are more appropriate, and the critical values for all criteria are less than 0.05, which means the differences of group cohesion are all significant. In addition, since the negative ranks are more than positive ranks in all criteria, the group cohesion in scheduling task is higher than that in rescheduling task.
- 28 -
Table 9 Task Interdependence and Group Cohesion
Wilcoxon Signed Ranks Test
N Mean Rank Sum of Ranks Z Asymp. Sig. (1-tailed) PGC Rescheduling – Scheduling Negative Ranks 16 12.00 192.00 -3.252 .001 Positive Ranks 4 4.50 18.00 Ties 0 Total 20 GEQ Rescheduling – Scheduling Negative Ranks 13 10.85 141.00 -2.428 .008 Positive Ranks 5 6.00 30.00 Ties 2 Total 20 TGC Rescheduling – Scheduling Negative Ranks 17 10.76 183.00 -2.914 .002 Positive Ranks 3 9.00 27.00 Ties 0 Total 20
Negative Ranks Rescheduling < PGC Scheduling Positive Ranks Rescheduling > PGC Scheduling Ties Rescheduling = PGC Scheduling
4.3.2. Communication Patterns and group cohesion
Table 10 Descriptive Statistics Group Cohesion
Group Statistics
Communication Patterns N Mean Std. Deviation Std. Error Mean PGC Computer Mediated 20 79.35 9.29 2.08
Face to Face 20 76.30 20.71 4.63
GEQ Computer Mediated 20 77.47 13.14 2.94
Face to Face 20 71.91 17.94 4.01
TGC Computer Mediated 20 78.22 10.72 2.40
Face to Face 20 73.67 18.34 4.10
- 29 -
face-to-face communication patterns in all criteria. The deviations are higher under face-to-face communication patterns than that under computer-mediated communication patterns. Hence, it is possible that communication patterns have great influence on group cohesion.
To better investigate the relationship between communication patterns and group cohesion, the normality tests are firstly applied to analyze the data under each criterion.
According to the normality test result, since the samples are not related，2 sample T test is suitable for testing the GEQ, and Wilcoxon sum rank test is more appropriate to explore the relationship here for PGC and TGC. In this case, single tailed tests are also used, and the critical values for all criteria are more than 0.05, which means the differences of group cohesion are not significant. Therefore, the communication patterns alone have no big influence on group cohesion.
Table 11 Normality Test for Communication Patterns--Group Cohesion
Tests of Normality Communication Patterns Shapiro-Wilk Normality Statistic df Sig. PGC
Computer Mediated .984 20 .975 yes Face to Face .803 20 .001 not
GEQ
Computer Mediated .938 20 .224 yes Face to Face .948 20 .344 yes
TGC
Computer Mediated .983 20 .964 yes Face to Face .883 20 .020 not
Table 12 Communication Patterns and group cohesion (GEQ)
2 sample T tests
Communication Patterns
Levene's Test for Equality of Variances
t-test for Equality of Means t df Sig. (1-tailed)
F Sig.
- 30 -
Table 13 Communication Patterns and group cohesion (PGC and TGC)
Wilcoxon Sum ranked tests
Communication Patterns N Mean Rank Sum of Ranks Mann-Whitney U Wilcoxon W Z Asymp. Sig. (1-tailed) PGC Computer Mediated 20 19.78 395.50 185.500 395.500 -.393 .347 Face to Face 20 21.23 424.50 Total 40 TGC Computer Mediated 20 21.08 421.50 188.500 398.500 -.311 .378 Face to Face 20 19.93 398.50 Total 40
However, the above findings have ignored the influence of differences in task interdependence. Hence, further analyses are conducted taking into account the different tasks performed.
Table 14 Descriptive Statistics
Group Statistics Group N Mea n Std. Deviation Std. Error Mean PGC
Computer Mediated Scheduling 10 83.89 10.26 3.24 Face to Face Scheduling 10 84.07 11.32 3.58 Computer Mediated Rescheduling 10 74.81 5.60 1.77 Face to Face Rescheduling 10 68.52 25.36 8.02
GEQ
Computer Mediated Scheduling 10 80.49 15.12 4.78 Face to Face Scheduling 10 76.67 16.00 5.06 Computer Mediated Rescheduling 10 74.44 10.76 3.40 Face to Face Rescheduling 10 67.16 19.32 6.11
TGC
Computer Mediated Scheduling 10 81.85 12.56 3.97 Face to Face Scheduling 10 79.63 13.43 4.25 Computer Mediated Rescheduling 10 74.59 7.47 2.36 Face to Face Rescheduling 10 67.70 21.22 6.71
- 31 -
under different communication patterns. Then, further statistical tests are applied to investigate the difference.
As illustrated in the result of normality test, Wilcoxon sum ranks tests are applied only for the comparison between PGC and TGC of computer mediated rescheduling and face to face rescheduling conditions. For the others, 2 sample tests are conducted. Again, single-tailed tests are used, but the critical values are all higher than 0.05. The findings indicate that only the differences under between each different criterion are not statistically significant. Therefore, it can be concluded that even the impact of task interdependence is considered, communication patterns still have no big influence on group cohesion.
Table 15 Normality Test for Group Cohesion
Tests of Normality
Shapiro-Wilk
Normality Statistic df Sig.
- 32 -
Table 16(1) Communication Patterns—Group Cohesion with regards to the same task
2 sample T tests
Group F Sig. t df Sig. (1-tailed)
PGC Computer Mediated Scheduling Equal variances assumed 0.68 0.42 -0.04 18.00 0.48 Face to Face Scheduling Equal variances not assumed -0.04 17.83 0.48 GEQ(1) Computer Mediated Scheduling Equal variances assumed 0.10 0.76 0.55 18.00 0.29 Face to Face Scheduling Equal variances not assumed 0.55 17.94 0.29 TGC Computer Mediated Scheduling Equal variances assumed 0.19 0.67 0.38 18.00 0.36 Face to Face Scheduling Equal variances not assumed 0.38 17.92 0.36 GEQ(2) Computer Mediated Rescheduling Equal variances assumed 2.59 0.13 1.04 18.00 0.16 Face to Face Rescheduling Equal variances not assumed 1.04 14.10 0.16
Table 16(2) Communication Patterns—Group Cohesion with regards to the same task
Wilcoxon sum Ranks
Group N Mean Rank Sum of Ranks Mann-Whitney U Wilcoxon W Z Asymp. Sig. (1-tailed) PGC Computer Mediated Rescheduling 10 9.65 96.50
41.500 96.500 -.645 .260 Face to Face Rescheduling 10 11.35 113.50
Total 20
TGC Computer Mediated Rescheduling 10 10.40 104.00
49.000 104.000 -.076 .470 Face to Face Rescheduling 10 10.60 106.00
Total 20
4.3.3. Communication Patterns and Perceived task interdependence (PTI) Table 17 Communication Patterns and PTI Descriptive Statistics
Descriptive Statistics
Communication Patterns N Mean Std. Deviation Std. Error Mean
PTI
Computer Mediated 20 75.00 21.30 4.76 Face to Face 20 79.64 19.70 4.41
According to the descriptive statistics, there is a slight difference between the average PTI and the deviation under different communication patterns.
- 33 -
Table 18 Normality Test for PTI
Tests of Normality Communication Pattern Shapiro-Wilk Normality Statistic df Sig. PTI
Computer Mediated .805 20 .001 not
Face to Face .786 20 .013 not
Table 19 Communication Patterns and PTI
Wilconxon sum ranks test
Communication Pattern N Mean Rank Sum of Ranks Mann-Whitney U Wilcoxon W Z Asymp. Sig. (2-tailed) PTI Computer Mediated 20 19.38 387.50
170.500 387.500 -.632 .528 Face to Face 20 21.63 432.50
Total 40
However, this finding might result from the impact of different tasks. Hence, advanced analysis is conducted to investigate the impact of different communication patterns on PTI under different tasks.
Table 20 Communication Patterns—PTI Descriptive Statistics
Group Statistics
Group N Mean Std.
Deviation Std. Error Mean
PTI
Computer Mediated Scheduling 10 86.43 7.10 2.25 Face to Face Scheduling 10 82.14 15.52 4.91 Computer Mediated Rescheduling 10 63.57 24.85 7.86 Face to Face Rescheduling 10 77.14 23.76 7.51
According to the descriptive statistics, the average PTI is slightly higher in computer-mediated scheduling than in face to face scheduling, while the average PTI is slightly higher in face-to-face communication rescheduling than in computer-mediated rescheduling. Hence, the communication patterns might have influence on perceived task interdependence.
- 34 -
Table 21 Normality Test for Group Cohesion
Tests of Normality
PTI
Shapiro-Wilk
Normality Statistic df Sig.
Computer Mediated Scheduling .774 10 .007 not Face to Face Scheduling .902 10 .228 yes Computer Mediated Rescheduling .892 10 .177 yes Face to Face Rescheduling .838 10 .042 not
Table 22Communication Patterns—PTI with regards to the same task
Wilcoxon sum ranks
Group N Mean Rank Sum of Ranks Mann-Whitn ey U Wilcoxon W Z Asymp. Sig. (2-tailed) PTI
Computer Mediated Scheduling 10 11.20 112.00
43.000 98.000 -.569 .569 Face to Face Scheduling 10 9.80 98.00
Total 20
PTI
Computer Mediated Rescheduling 10 9.00 90.00
35.000 90.000 -1.156 .248 Face to Face Rescheduling 10 12.00 120.00
Total 20
According to the critical value depicted above, the differences are not significant in both situations. Therefore, communication patterns have no significant influence on perceived task interdependence.
4.3.4. Interaction Effects
To test the interaction effects, various methods were used. Firstly, the data was categorized into four different groups, namely computer-mediated scheduling, computer-mediated rescheduling, face-to-face scheduling, and face-to-face rescheduling. The normality test results are the same as depicted in Table 11.
- 35 -
applied to investigate the GEQ.
The related statistic description and normality tests results are depicted separately in Table 14 and Table 15.
Table 23 Interactive Effect (1)
Kruskal Wallis Test
Group N
Mean Rank
Chi-Square df Asymp. Sig.
PGC
Computer Mediated Scheduling 10 25.90
6.944 3 .074 Computer Mediated Rescheduling 10 14.10
Face to Face Scheduling 10 24.45 Face to Face Rescheduling 10 17.55
Total 40
TGC
Computer Mediated Scheduling 10 26.35
4.807 3 .186 Computer Mediated Rescheduling 10 16.60
Face to Face Scheduling 10 22.20 Face to Face Rescheduling 10 16.85
Total 40
Table 24Interactive Effect (2)
One way ANOVA
N Mean
Std. Deviation
Std. Error
Sum of Squares df Mean Square F Sig.
Between 943.454 3 314.485 1.292 .292 Computer Mediated Scheduling 10 80.49 15.12 4.78 Computer Mediated Rescheduling 10 74.44 10.76 3.40
Face to Face Scheduling 10 76.67 16.00 5.06
Within 8762.384 36 243.400 Face to Face Rescheduling 10 67.16 19.32 6.11
- 36 -
According to the above result, there is no significant difference in Group Cohesion. But when taking the mean rank and the mean value into account, it seems Computer Mediated Scheduling has the highest group cohesion in each criterion.
By far the interaction effect has not been proved yet. Detailed analyses were conducted based on different independent variables. Two nominal variables, namely the task interdependence and communication patterns, were involved in this analysis. Here attached the figure showing interaction effect.
Figure 4 Interactive effect Plot (PGC)
- 37 -
Figure 6 Interactive effect Plot (PTC)
As illustrated in all three graphs, it is obvious that the group cohesion is higher in scheduling than that in rescheduling, and it is higher when communicating through computer media rather than face to face. Moreover, the difference of group cohesion between scheduling and rescheduling tasks is slightly lower in computer-mediated mode than that in face to face mode. Therefore, it is possible that the slight difference is caused by the interaction effect.
To test if the interactive effect exists, the difference mentioned above is firstly calculated according to the following formula,
D1=
D2=
- 38 -
Table 25Normality Test for Differences
Tests of Normality Shapiro-Wilk Normality Statistic df Sig. D1PGC .957 10 .754 yes D2PGC .752 10 .004 not D1GEQ .953 10 .700 yes D2GEQ .884 10 .145 yes D1TGC .934 10 .484 yes D2TGC .821 10 .026 not
Table 26Interactive Effect-- Group Cohesion Differences
2 sample T test
GEQ N Mean Std. Deviation Std. Error Mean
Computer Mediated 10 6.19048 12.905065 4.080940
Face to Face 10 10.00000 15.059395 4.762199
F Sig. t df Sig. (2-tailed)
Equal variances assumed .011 .918 -.607 18 .551
Equal variances not assumed -.607 17.587 .551
Table 27 Interactive Effect—Group Cohesion Differences
Wilcoxon sum Ranks
- 39 -
5. Discussion
5.1. Discussion of the results
Previous research shows that communication patterns, task interdependence and group cohesion are crucial factors with regards to group process and group work outcome. However, the interrelationship still remains unclear. More specifically, the role of task interdependence is confusing, whether it serves as a moderator or mediator. In this study, by conducting a controlled experiment, the relationships among the above three factors were studied. The findings demonstrate the interrelationships in several ways.
Firstly, the results indicate that task interdependence has a direct impact on group cohesion. A higher level of task interdependence can lead to higher group cohesion. Meanwhile, under the same level of task interdependence, communication patterns make no difference on group cohesion. Moreover, communication patterns do not have significant impact on task interdependence. Hence, hypothesis 1 should be accepted, while hypothesis 3 and hypothesis 4 should be both rejected．In line with
Vijfeijken & Kleingeld (2002), in a high task interdependence condition, more information exchanges are necessary to fulfill the group task. In addition, the more time people spend together, the stronger their group cohesion becomes (Levine & Moreland 1990). By exchanging more information, more time is needed to communicate, thus resulting in higher group cohesion. Therefore, task interdependence is not a mediator in the relation between communication patterns and group cohesion.
- 40 -
in computer-mediated scheduling situation is the highest in the four conditions, the difference is not statistically significant.
To summarize, communication patterns have no impact on group cohesion, while the level of perceived task interdependence directly affects the degree of group cohesion.
5.2. Limitation
The above findings are inconsistent with Yoo & Alavi(2001) that communication medium and group cohesion are related. Since there is a slight difference in the mean group cohesion for different conditions, it is possible that experiment results are biased. There are four main reasons that may lead to the biased findings.
Firstly, the participants are biased. On the one hand, they have different level of knowledge of the experiment task. On the other hand, when it comes to communication, it can be influenced by the different personality of participants.
Secondly, the sample size is too small that the variation is not so big within different groups. According to the descriptive statistic data, the group cohesion scores are slightly different between different conditions. However, when the statistical tests are conducted, the differences are not significant. Due to the lack of data, the standard deviation of group cohesion scores for each experimental group big, If more data are available, it is possible that different findings can be gained.
Moreover, the experimental task, the scheduling task, is a problem solving task. The findings can be different when a decision making task is applied. Even the time to perform the group tasks have an impact on the degree of group cohesion as the longer it takes the higher the group cohesion. (Levine & Moreland 1990)
- 41 -
5.3. Contribution to Theory and Managerial Relevance
This study contributes to theory and practice in several ways.Existing literature shows that all three factors studied are vital considering group work processes and outcomes. Earlier research only indicates that communication medium and group task norm are all related to group cohesion and performance. There is a literature gap on the interrelationship among communication patterns, task interdependence and group cohesion. This study contributes to this gap by making the interrelationships more clear. The communication patterns alone have no significant impact on group cohesion and perceived task interdependence. Group cohesion is directly positively affected by perceived task interdependence.
This finding has managerial relevance as well. On the one hand, managers might be able to decide a proper communication pattern to conduct the group task, for instance, whether to have daily face-to-face meetings or daily on-line meetings to assign group tasks. According to the findings here, it is not necessary to conduct a face-to-face meeting when communication patterns have no big impact on group cohesion, especially when computer-mediated pattern is more convenient. On the other hands, when there is groupthink in a work group, it is possible to reduce group cohesion so as to curtail the impact of groupthink. Since task interdependence positively affects group cohesion, the group cohesion in a work group can be reduced by assigning less interdependent tasks to work group members.
5.4. Discussion of future directions
- 42 -
Levine & Moreland （1990), the more time people spend together, the stronger their cohesion becomes. As a result, the former work group may have a higher degree of group cohesion than the latter one. Therefore, communication frequency can be positively related to group cohesion, and this should be studied.
Secondly, as mentioned before, the different task type might lead to different findings. Experiment with the decision making task should be conducted to see if there is any difference.
- 43 -
6. Conclusion
- 44 -
7. Reference
Axelord, R., 1984. The Evolution of Cooperation. In Basic Books, New York. 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(6), pp.1173–1182.
Bendoly, E. et al., 2009. Bodies of Knowledge for Research in Behavioral Operations. Production and Operations Management, 19(4), pp.434–452. Bendoly, E., Bachrach, D.G. & Powell, B., 2008. The Role of Operational Interdependence and Supervisory Experience on Management Assessments of Resource Planning Systems. Production and Operations Management, 17(1), pp.93–106.
Bendoly, E., Donohue, K. & Schultz, K.L., 2006. Behavior in operations management: Assessing recent findings and revisiting old assumptions. Journal
of Operations Management, 24(6), pp.737–752.
Bernthal, P.R. & Insko, C. a., 1993. Cohesiveness without Groupthink: The Interactive Effects of Social and Task Cohesion. Group & Organization
Management, 18(1), pp.66–87.
Bishop, J.W. & Dow Scott, K., 2000. An examination of organizational and team commitment in a self-directed team environment. Journal of Applied
Psychology, 85(3), pp.439–450.
- 45 -
Bollen, K.A. & Hoyle, R.H., 1990. Perceived Cohesion : A Conceptual and Empirical Exandnation * at Chapel. Social Force, 69(2), pp.479–504.
Campion, M.A. & Medsker, G.J., 1993. Relations Between Work Group Characteristics and Effectiveness: Implications for Designing Effective Work Groups. Personnel Psychology, 46, pp.823–849.
Carless, S. a. & De Paola, C., 2000. The Measurement of Cohesion in Work Teams. Small Group Research, 31(1), pp.71–88.
Carron, a. V. & Spink, K.S., 1995. The Group Size-Cohesion Relationship in Minimal Groups. Small Group Research, 26(1), pp.86–105.
Carron, A. V. & Brawley, L. R., 2000. Cohesion: Conceptual and Measurement Issues. Small Group Research, 31(1), pp.89–106.
Chang, A. & Bordia, P., 2001. A multidimensional approach to the group cohesion-group performance relationship. Small Group Research.
Chin, W.W. et al., 1999. Perceived Cohesion in Small Groups: Adapting and Testing the Perceived Cohesion Scale in a Small-Group Setting. Small Group
Research, 30(6), pp.751–766.
Delbecq, A., 1974. The effectiveness of nominal, delphi, and interacting group decision making processes. Academy of Management Journal, 17(4), p.337.
Evans, C.R. & Dion, K.L., 1991. Group Cohesion and Performance: A Meta-Analysis. Small Group Research, 22(2), pp.175–186.
Evans, N.J. & Jarvis, P. a., 1980. Group Cohesion: A Review and Reevaluation.
Small Group Research, 11(4), pp.359–370. Eys, M. a et al., 2007. Item wording
- 46 -
Furumo, K. & Pearson, J., 2006. An empirical investigation of how trust, cohesion, and performance vary in virtual and face-to-face teams. … , 2006.
HICSS’06. Proceedings of the …, 00(C), pp.1–10.
Gladstein, D.L., 1984. Groups in Context : A Model of Task Group Effectiveness.
Administrative Science Quarterly, 29(4), pp.499–517.
Goodman, P., Ravlin, E. & Argote, L., 1986. Current thinking about groups: Setting the stage for new ideas.
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), pp.497–520.
Guzzo, R. a & Dickson, M.W., 1996. Teams in organizations: recent research on performance and effectiveness. Annual review of psychology, 47, pp.307–38. Harrison, D.A., Price, K.H. & Bell, M.P., 1998. Beyond Relational Demography : Time and the Effects of Surface- and Deep-Level Diversity on Work Group Bell Reviewed work ( s ): Source : The Academy of Management Journal , Vol. The
Academy of Management Journal, 41(1), pp.96–107.
Hawkey, K. et al., The Proximity Factor : Impact of Distance on Co-located Collaboration. , pp.31–40.
Hogg, M.A., 1992. The social psychology of group cohesiveness: From
Attraction to Social Identity,
- 47 -
De Jong, S.B., Van der Vegt, G.S. & Molleman, E., 2007. The relationships among asymmetry in task dependence, perceived helping behavior, and trust.
The Journal of applied psychology, 92(6), pp.1625–37
Kennedy, D.M., Vozdolska, R.R. & McComb, S. a., 2010. Team Decision Making in Computer-Supported Cooperative Work: How Initial Computer-Mediated or Face-to-Face Meetings Set the Stage for Later Outcomes*. Decision Sciences, 41(4), pp.933–954.
Keyton, J. & Springston, J., 1990. Redefining Cohesiveness in Groups. Small
Group Research, 21(2), pp.234–254.
Kiesler, S., Siegel, J. & McGuire, T.W., 1984. Social psychological aspects of computer-mediated communication. American Psychologist, 39(10), pp.1123–1134.
Langfred, C.W., 1998. Is Group Cohesiveness a Double-Edged Sword?: An Investigation of the Effects of Cohesiveness on Performance. Small Group
Research, 29(1), pp.124–143.
Levine, J. & Moreland, R., 1990. Progress in small group research. Annual
review of psychology, 41, pp.585–634. Available at:
Mullen, B. et al., 1994. Group Cohesiveness and Quality of Decision Making: An Integration of Tests of the Groupthink Hypothesis. Small Group Research, 25(2), pp.189–204.
- 48 -
Preston, C.C. & Colman, a M., 2000. Optimal number of response categories in rating scales: reliability, validity, discriminating power, and respondent preferences. Acta psychologica, 104(1), pp.1–15.
Rajeev, S. & Yetton, P., 2003. The Contingent Effects of Management Support and Task Interdependence on Successful Information Systems Implementation.
MIS Quarterly, 27(4), pp.533–556.
Ramamoorthy, N. & Flood, P.C., 2004. Individualism/Collectivism, Perceived Task Interdependence and Teamwork Attitudes among Irish Blue-Collar Employees: a Test of the Main and Moderating Effects? Human Relations, 57(3), pp.347–366.
Rhoads, M., 2010. Face-to-Face and Computer-Mediated Communication: What Does Theory Tell Us and What Have We Learned so Far? Journal of Planning
Literature, 25(2), pp.111–122.
Rico, R. et al., 2009. The joint relationships of communication behaviors and task interdependence on trust building and change in virtual project teams. Social
Science Information, 48(2), pp.229–255.
Rogelberg, S.G. & Rumery, S.M., 1996. Gender Diversity, Team Decision Quality, Time on Task, and Interpersonal Cohesion. Small Group Research, 27(1), pp.79–90.
Saavedra, R., Barley, P.C. & Dyne, L. Van, 1993. Complex Interdependence in Task-Performing Groups. Journal of Applied Psychology, 78(1), pp.61–72. Siebold, G.L., 1985. The Evolution of the Measurement of Cohesion. Military
- 49 -
Snoo, C. De & Wezel, W. Van, 2011. Coordination and Task Interdependence during Schedule Adaption. Human Factors and Ergonomics in Manufacturing &
Service Industries, 00(0), pp.1–13.
Stokes, J.P., 1983. Components of Group Cohesion: Intermember Attraction, Instrumental Value, and Risk Taking. Small Group Research, 14(2), pp.163–173. Straus, S.G. & McGrath, J.E., 1994. Does the medium matter? the interaction of task type and etchnology on group performance and memeber reactions. Journal
of Applied Psychology, 79(1), pp.87–97.
Summers, I., Coffelt, T. & Horton, R.E., 1988. Work-Group Cohesion.
Psychological Reports, 63(2), pp.627–636.
Vegt, G. Van Der & Vliert, E. Van De, 2002. Intragroup interdependence and effectiveness: Review and proposed directions for theory and practice. Journal of
Managerial Psychology, 17(1), pp.50–67.
Van de Ven, A., Delbecq, A. & Jr, R.K., 1976. Determinants of coordination modes within organizations. American sociological review, 41(2), pp.322–338. Vijfeijken, H. Van & Kleingeld, A., 2002. Task complexity and task , goal , and reward interdependence in group performance management : A prescriptive model. European Journal of Work and Organizational Psychology, 11(3), pp.363–383.
Wageman, R., 1995. Interdependence and Group Effectiveness. Administrative
Science Quarterly, 40(1), pp.145–180.
Wageman, R. & Baker, G., 1997. Incentives and cooperation : the joint e € ects of task and reward interdependence on group performance. Journal of
- 50 -
