Rijksuniversiteit Groningen
Evaluators: Karin Prins & Ben Emans August 2005
Student-Team Performance Ratings as Influenced by Team Conscientiousness Levels
Master’s Thesis Human Resource Management Clarity Wolff
S1362755
Table of Contents
Abstract………...3
Introduction………...3
The Big Five Personality Factors………..4
Background of the Study………...5
Unitary Tasks……….6
Team Tasks……….7
Research Questions………8
Method………....9
Respondents……….9
Procedure……….9
Measures………..10
Results……….10
Discussion……….13
Reference List………...16
Tables Table 1………11
Zero-Order Correlations Table 2………..12
Correlation among Each of the Big Five Personality Traits with Controlling for Team Size Table 3………..13
Mediating Effects of Communication on the Conscientiousness-Performance Relationship
Student-Team Performance Ratings as Influenced by Team Conscientiousness Levels
Abstract1
This study examines how the Big Five Personality Traits impact performance of teams. Our sample consisted of 11 student teams who voluntarily completed the Organizational Big Five Inventory (OBFI), which is used to measure personality characteristics. We analyzed three hypotheses designed to test for: relationships between mean level of conscientiousness and performance, homogeneity of conscientiousness in a team and its performance, and we tested for moderating effects of communication on the conscientiousness-performance relationship.
Our first and third hypotheses were supported; homogeneity of conscientiousness was not shown to be significant with relation to team performance. We demonstrated a positive relationship between team personality, most specifically conscientiousness, and performance.
This is an area of research important to the success of the growing number of organizations which rely more now than ever upon the use of teams.
Introduction
Several recent trends are creating additional interpersonal contact for employees since they must interact and communicate more with co-workers and customers alike. Initially, we have seen a dramatic increase in the size of the service sector. More and more frequently the customer is involved in the process of production and delivery of a service (Mount, Barrick, Stewart 1998). Another trend we are seeing is that many businesses are organizing their employees into teams and assigning them to projects. Deeprose (1995) highlights several example organizations as members of an increasing number of companies that have made a reallocation of supervisory responsibilities from managers to teams. This is an example of teams being given new responsibilities. Subsequently if such teams are incapable of properly working together and communicating to carry out their group tasks, the entire organization may feel the repercussions of this failure.
The abundance of self-managed teams in the work place creates new questions about the psychology of group performance (Barry & Stewart 1997). In Human Resource Management, a significant amount of time is spent on understanding the skills and abilities of the people who are working for or soliciting work from the organization. It is important to the overall success of any organization that its employers properly utilize the skills and abilities available to them. Increasing in significance for individuals, managers, and organizations is having an understanding of why some teams are effective while others are not. Knowing what actions can and should be taken to prevent team failure is beneficial to the success of the individuals, managers, and organizations alike.
Considerable research has been done with the goal of better understanding personality and its effect on employee capacity and motivation. The subject of personality has been studied at length; according to Hogan (1991), the developments in personality psychology in the last twenty years proposes that people can be characterized in terms of their enduring dispositional qualities and that this information can be used in ways that have significance for employee development as well as organizational effectiveness. For some time now theorists,
We wish to thank Karin Prins & Ben Emans for guidance, assistance and a critical eye. To Dan and Hayley Eveleth a warm thanks for your support and encouragement. We would also like to extend our appreciation to
scholars, and employers have understood that individual personality characteristics can be linked to employee performance. Nevertheless, the study of how individual personalities come together effectively or ineffectively in a group or team setting is comparatively just beginning to gather attention from researchers. “Despite an increasing number of organizations structuring work through the use of teams, we know relatively little about how diverse individuals comprising a team influence team outcomes” (Mohammed, Mathieu &
Bartlett, 2002).
Glassman (2000) has drawn attention to the ways that each member of a sports team contributes certain qualities to create a winning formula. She also discusses how in an ideal world, the personalities of the players would provide a perfect fit to the ‘role’ they play on the team and that performance suffers when team members do not play their part. In this study, we look at the roles personality plays in influencing team performance in an academic setting. Moreover, this study seeks to show that personality measures, such as the Organizational Big Five Inventory (OBFI), can serve as objective tools to determine the personalities of potential team members. It is crucial for the continued development of team- personality theory that multiple studies be conducted with varying samples to widen the foundation upon which theory can be built.
While our academic exploration serves to expand overall understanding of how the Big Five personality factors relate to team performance it concurrently focuses on one of the Big Five factors in particular, conscientiousness. With our research we aimed to identify personality characteristics that could be attributed to student teams that produce high levels of performance. Our focus was placed on the influence of conscientiousness on a student teams’
performance satisfaction.
The Big Five Personality Factors
In literature on personality, the Big Five personality factors have recently become widely accepted. The factors have been used in studies covering an incredible array of situations;
numerous samples have been studied using many different instruments for measuring these Big Five traits. Big Five Assessment is a compilation of Big Five inventories and questionnaires, Big Five adjective scales, and other Big Five associated instruments. These have been conglomerated in one book to provide an in depth look at the Big Five (De Raad and Perugini 2002).
The Big Five traits have been used across various questionnaires and inventories; in this study we use the Organizational Big Five Inventory (OBFI) (van Oudenhoven et al., 2003).
Our English version of the Dutch original is included in Appendix 1. The OBFI has not been as widely used as others such as the Five Factor Model or the Personal Characteristics Inventory. However, this inventory has much strength. The short length of the OBFI makes it an attractive inventory to use. Another benefit of the OBFI is that it was designed in such a way as to be applicable at various levels of an organization. The OBFI measures personality by surveying how individuals interact with other members of their team, and how they perceive themselves to be viewed by others.
Considering there has been a long development of personality studies, there are multiple schemas to refer to when outlining the definitions of personality characteristics. Of the numerous Big Five instruments, not all of them use precisely the same terms to label the
scales; we found our terminology to be the most clearly understandable of the various terms in circulation. To prevent confusion an outline of the terminology we will use is in order.
Agreeableness
Highly agreeable people are often seen as tolerant, friendly and willing to make compromises. These people will often be trustworthy and compliant.
Conscientiousness
This characteristic can be seen in individuals who are dependable, efficient, organized, and hardworking tend to be better performers than people not possessing these behaviors. The majority of evidence shows conscientiousness as an important correlating factor to job performance (Mount, Barrick & Strauss, 1999).
Emotional Stability
The emotionally stable person can take his or her mind off of problems and is thus able to keep working even in non-ideal circumstances. Emotionally stable people will typically score low for traits such as anxiety and impulsiveness.
Extraversion
This dimension of the Big Five is associated with people who are typically outgoing, talkative and sociable. When people score high on the extraversion scale they are likely to be seen as warm, positive and, assertive types.
Openness to Experience
In Hendricks, Hofstee and De Raad (2002, p. 89) they refer to people with this characteristic as having the ability to easily link facts together, wanting to form own opinions, and analyzing problems. People who are highly open to experience may also often have creative new ideas or be prone to flights of the imagination.
Background of the Study
With this brief overview of the Big Five factors, we are better prepared to expand on the aim of this study. Witt and Ferris (2003) state that conscientiousness is possibly the most consistent predictor of job performance. With this in mind our study aims to identify personality characteristics that can be attributed to student teams with high levels of performance and primarily focuses on the Big Five personality characteristic, conscientiousness.
Many of the studies that have been conducted with regard to group personality and team outcomes have been conducted in laboratory settings. Relatively little has been done to understand the relationship that personality has on performance of actual work teams completing production tasks (Barrick et al., 1998). Due to the difficulties of finding samples with a large number of teams this study will consider hypotheses and methods used in several other studies conducted on team personality (Barrick et al. 1998, van Vianen and De Dreu, 2001).
The current study was inspired by the work of Barrick et al. (1998) and van Vianen & De Dreu (2001). Our intention with this study is not to reproduce their work nor to further test all of their hypotheses, but rather to supplement the breadth of knowledge and understanding regarding the effects personality may or may not have on groups in various working environments. This study may offer some results comparable to results found in these other studies.
Barrick et al. (1998) conducted a study which connected ability and personality to team process and effectiveness. Their study was composed of 51 work teams (652 employees) who were examined using the five-factor model to measure personality and its relationship to performance. Their study showed that teams with higher levels of general mental ability, conscientiousness, agreeableness, extraversion, and emotional stability earned higher team performance ratings from their supervisors. Teams in their study that were higher in general mental ability, extraversion, and emotional stability earned higher supervisor ratings for team viability. Their results also show associations between extraversion and emotional stability through social cohesion.
Van Vianen & De Dreu (2001) continued the research of Barrick et al (1998). They compared the sample from Barrick et al. to their own two sets of samples. One set was comprised of 24 drilling teams in the US and the other sample was of 25 student teams from the Netherlands.
In the effort to further the usefulness of results from Barrick et al. their study considered the relationship between personality composition and task cohesion, which they considered to be a better predictor of performance than social cohesion. Their results partially confirmed the results from Barrick et al. with regard to personality composition, cohesion, and team performance. Significant relationships were found between social cohesion, task cohesion, and performance; yet the measures of cohesion did not mediate relationships between personality composition and team performance.
Unitary Tasks
In the above-mentioned studies as in ours, we have considered what different tasks are expected of the student team members. It is important to determine what sorts of actions lead to desired results when measuring team performance; since there are different types of tasks, requiring different means of performance to be successfully completed. Repeatedly researchers have gauged the mean scores per team for their measures, and as a result have taken the position the more, the better. Yet, according to the work of Steiner (1972), there are both divisible and unitary tasks. Divisible tasks can be easily broken down into subtasks whereas unitary tasks cannot. “Two persons who read alternate sentences of this paragraph will not gain a better understanding of its content than will a single reader.” (Steiner, 1972 p.15).
The work required of our sample best fits the definition of unitary tasks. There are four different types of unitary tasks; disjunctive, conjunctive, additive, and discretionary. Here we will briefly explain these different tasks, and later we will discuss which of these we will specifically focus on with our sample and why.
Disjunctive tasks are ones where the performance of the group is actually determined by the performance of one team-member. When a team must find a solution to a problem that has several alternatives, this can be seen as disjunctive because only one answer can be given, and therefore one team member could make this decision. To measure disjunctive tasks it is useful to measure the maximum individual score per team since the best team member will establish the team outcome.
With Conjunctive tasks, the team is only as strong as its weakest member. The team will only perform as well as the individual who performs the least well. An example of such a task is a race. If a team runs a race and the whole team has to cross the finish line to get their ranking,
then the team will only do as well as their slowest runner. Having considered this, to measure the results of such a task one should use the results from the lowest scoring team member.
Additive tasks use the sum of each individual’s effort. This is added to create a total score for the team. Unlike the two previous types of tasks, where either the best or worst performance is assigned to all team members, additive tasks use the mean level per team (the the more, the better view) since all members contribute to the total. This can be clarified with the task of moving a heavy object; the team would work together to achieve the goal.
When assigned to a discretionary task, teams are given the free reign to determine how they will combine their individual contributions. The group can assign different weights to each individual if they so choose. This type of task is frequently seen when groups are required to do forecasting or estimations. To best measure a team’s ability with discretionary tasks variance of scores per team should be used.
Team Tasks
Considering the great amount of time the student teams in this study spend working together on various projects it is arguable that the students encounter all of the different tasks outlined in the section on unitary tasks. It would therefore be justifiable to use any of the forms of unitary tasks to measure performance. For the purposes of this study, we have chosen to focus on two of the four, based on their relevance and their importance in satisfying the requirements of the business program. These measures will then be calculated following Steiner (1972) and his work on unitary tasks.
The students in these teams are interdependent in that they are required to work together to achieve common goals. While there are independent tasks such as frequent examinations, the teams are also required to complete several team projects and presentations. The work the student teams do is conglomerated to create one final product, at which point it is no longer clear who was responsible for which tasks. These projects and presentations are assigned and evaluated as a whole. It is fully at the discretion of the team members to determine who will do what and how to divide the work. To this extent the tasks are additive in nature and we will accordingly measure the team compositions using mean scores per team, as explained in Steiner (1972), from the inventory.
Due to the complexity and size of the assignments given to the student teams it is not possible for the students to complete the project without proper cooperation, and without making the best use of each individual’s strengths. For example, if the students divide sections of the project up into individual tasks, each member of the team should be assigned tasks based on their ability to complete them. These large projects have multiple aspects, specifically the students are required to write a paper and give a presentation on their findings. The project is designed to improve students’ ability to work with and understand things such as financial documents, marketing schemas, and productions methods. Often the students comprising the teams are majoring in different areas of business (such as, management, marketing, finance, accounting, or economics) this implies that not all team members are equally equipped with the same set of skills (which were gained in other courses specifically required by different majors). This illustrates that some of the team tasks may best be viewed as discretionary tasks, for which we will calculate the variance of team scores.
Research questions
Having provided a brief explanation of these different types of tasks and of the sample teams we can now determine the hypotheses to be measured by this study. The taxonomy provided by Steiner (1972) helps us to properly test our hypotheses by using the most appropriate measurement methods.
Across several studies conscientiousness has been found to be a valid predictor of individual performance (Barrick & Mount, 1991; Hough et al, 1990). Since it has been shown that conscientiousness can influence individual performance in the above mentioned studies; we have chosen to focus predominately on this personality characteristic.
When a team is being rated based upon additive or discretionary tasks each individual in the team will have an impact on the score of the group. From this, we have made the prediction that conscientiousness will influence the performance of our sample at the group level based upon its individual level influence. We are interested in testing whether or not teams can have higher performance based upon their team conscientiousness levels.
The work of van Vianen & De Dreu (2001) and Barrick et al. (1998) has been used as a springboard for the development of our own hypotheses. Barrick et al. (1998) showed that work teams with higher mean levels of conscientiousness received higher team performance scores from supervisors. In addition, minimum levels of conscientiousness were significantly related to team performance ratings in the study done by van Vianen & De Dreu (2001).
Based upon prior support for the link between conscientiousness and performance ratings we predict that the mean level of individual conscientiousness will be positively related to the work teams’ ratings of team performance (hypothesis 1).
In this study we will also measure how the mean levels of the other four personality characteristics measured (extraversion, agreeableness, emotional stability, and openness to experience) are related to the team performance ratings.
Van Vianen & De Dreu (2001) predicted that groups having more homogeneous scores of conscientiousness would have higher task cohesion. “Individuals low in conscientiousness share the tendency to be disorganized and approach the task in a less structured way. When a team consists of both high and low conscientious members, there may be substantial disagreement and irritation about how to approach the task, and what to achieve.” (p.100).
Their results were in the predicted direction but were not statistically significant.
From this we question whether our sample might be able to offer a statistically significant result where theirs did not. We will test if work teams having similar scores for conscientiousness also have high ratings of team performance (hypothesis 2).
Along with this we will also measure the variance of the other four items (extraversion, agreeableness, emotional stability, and openness to experience) in relation to team performance.
In addition to considering conscientiousness and other elements of personality we will investigate the influence that a team’s ability to communicate will have on team performance.
In our inventory we asked participants what their level of satisfaction was regarding communication in their team. We have chosen to search for mediating effects of communication because of its appeal to logic; it seems logical to presume that a team with higher levels of communication will in turn be better equipped to perform at a high level.
Huseman et al. stated that “the relationship between communication and performance, while
intuitively appealing, is decidedly unclear.” (1980, p. 178). We would like to explore this and thus predict that highly conscientious teams, which also score highly for communication ratings, will have higher performance ratings than conscientious teams lacking in communication skills (hypothesis 3).
Method
Respondents
The original sample for this study was comprised of 80 upper-division university students participating in an intensive business program at the University of Idaho. The program length is one scholastic year. The students are taught by a team of five instructors, each having a different specialization within the field of business.
At the beginning of this program, the students are taught about teams for several lectures and are then instructed to organize themselves into teams. These teams are required to work together to complete several projects and presentations. In addition to the teamwork, for which they receive grades as a team, there are also numerous individual tasks and examinations that each student must complete. The students are required to attend lectures for this program nine hours a week. Every third week the students have an examination in addition to the lecture hours. Beyond this, the teams meet on their own time to complete their research projects, write papers, and prepare and practice their presentations for this course.
This program serves to teach business students, among other things, what it means to be professional. Business attire is required for presentations and for the majority of students this is the first time they have ever been required to present a speech without the assistance of note cards.
Of the 80 respondents (73% male and 27% female) we gathered results from 19 teams, of which 11 we considered complete teams. All teams had either five or six members, the teams that we used in our research, had 4, 5 or 6 members (the results included five 4-person teams, three 5-person teams, and three 6-person teams). Consequently we used 11 teams in our study for a total of 53 respondents.
Procedure
The Organizational Big Five Inventory (OBFI) is unique from many other questionnaires and inventories which are based upon the Big Five Personality Characteristics in that it can be applied at different levels within an organization (van Oudenhoven et al., 2003). The terminology in this inventory has been specifically tailored to be grammatically applicable to various levels in an organization.
To use this inventory for our study we first translated it with a bilingual panel to English from its original version in Dutch. Some believe it necessary to use back translating, in this case that would suggest translating the English version of the OBFI back into Dutch to control for accuracy. “Questionnaires are notoriously difficult to translate into other languages. When items need to be adapted for each language version, comparability of results among versions becomes questionable.” (Hendriks, Hofstee, De Raad, 2002). Opposing this view, Hofstede (1984) has argued that if a bilingual panel translates a questionnaire and if the readers are familiar with the terminology that is being translated it may be just as effective as other translations, which have made use of back translating.
Output and process measures (personality, performance, communication) were collected from each participant. An instructor of this course distributed the inventory to the students and offered the respondents time in class to complete the OBFI if they were willing to. All of the respondents voluntarily and anonymously completed the English version of the OBFI. It may be important to note that the inventories were handed out in the last week of the program, results were known for some projects, yet final grades were not yet known to the students.
Measures
This inventory is a 36-item inventory designed to measure personality at individual, team, or organization-wide levels. The five scales (agreeableness, conscientiousness, emotional stability, extraversion and openness to experience), each contained 6 items. Based on results from a factor analysis and coefficient alpha we removed one question from the agreeableness scale and one question from emotional stability scale leaving those two scales with 5 items and the remaining three with 6 items, resulting in a total of 34-items. All five scales showed reasonable reliabilities ranging from α= .64 − α= .73.
Each question was introduced with ‘In general I…’; we then asked the participants to chose the answer that best matched their opinions of themselves on 5-point scales ranging from 1=
Strongly Disagree to 5= Strongly Agree. Examples of the scales are: like to be in the spotlight (extraversion), do my best to help others (agreeableness), focus on planning (conscientiousness), stay calm in difficult situations (emotional stability), and try new ideas (openness to experience).
The results from a study testing the quality of single-item measures provide support for the use of single-item measures used to measure job satisfaction. While single item measures may have some limitations, research done by Wanous, Reichers & Hudy (1997) showed single-item measures to have no effect on results of a meta-analysis, whereas when scales were measured differently it consequently produced different results. “Because of this difference, it seems reasonable to conclude that the single-item measures are more robust than the scale measures of overall job satisfaction.” With this in mind we used single-item estimations to measure performance satisfaction as well as communication. These questions were introduced with ‘In general how satisfactory was your…’. For our purposes we used the measures of how each individual viewed the team’s ability to perform and communicate.
Results
The results that we gathered provided us the chance to work towards our aim of further developing the implications of existing relationships between personality and performance.
The data were initially analyzed by calculating zero-order correlations, for this we calculated the mean, variance, minimum, and maximum between the five personality scales (see Table 1). The correlations are the highest among different methods of operationalizing the same trait. The average of these correlations (using absolute values) using the four different operationalizing methods (mean, variance, maximum, and minimum) as explained by Steiner (1972) was .53 for extraversion, .52 for agreeableness, .51 for conscientiousness, .53 for emotional stability, and .42 for openness to experience. The average of these correlations across the five personality scales was .50.
The average correlation for same methods-different traits (i.e. all mean scales correlated with the mean measures from other scales) was much smaller, namely .26. The measure of
different traits-different methods (i.e. mean for extraversion correlated with variance, minimum, and maximum for the other scales) the correlation was also .26.
TABLE 1 Zero-Order Correlations for Team Personality Variables Operationalized Using the Mean, Median, Mode, Variance, Minimum and Maximum Methods (N=11)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Extrave rsion 1 Mean
2 Variance -.16
3 Minimum ,76** -.70
4 Maximum ,84** .33 .41
Agre eable ne ss
5 Mean -.13 -.09 .02 -.23
6 Variance .14 .51 -.24 .33 -,56*
7 Minimum -.11 -.12 .04 -.24 ,93** -,67*
8 Maximum -.13 ,64* -.43 .19 .01 ,69** -.23
Consie ntiousne ss
9 Mean .23 .07 .17 .19 .08 .35 -.06 .42
10 Variance .20 .40 .05 .52 -.28 ,59* -.42 ,65* .15
11 Minimum .06 -.30 .14 -.25 .09 -.06 .12 -.23 ,63* -,61*
12 Maximum .13 .14 .15 .22 -.04 .38 -.19 .49 ,89** .47 .32
Emotional Stability
13 Mean -.16 .23 -.38 -.01 -.31 .46 -,58* .50 .06 .21 -.14 .05
14 Variance -.27 .36 -.42 -.13 .51 .06 .42 .41 .22 .00 -.02 .14 .15
15 Minimum -.20 -.06 -.19 -.14 -,53* .18 -,67* .07 -.24 .05 -.14 -.18 ,72** -.51 16 Maximum -.41 .43 -0,62* -.13 -.07 .28 -.30 .51 -.10 .24 -.36 -.02 ,86** .48 .46 Autonomy
17 Mean .30 .05 .07 .39 -,77** .18 -,74** -.19 -.25 .05 -.14 -.22 .42 -,60* ,71** .16 18 Variance .29 -.32 .45 .31 -.26 -.13 -.23 -.16 -.38 .38 -,64* -.16 -.05 -.12 .02 .00 .24 19 Minimum .08 .36 -.28 .20 -,60* .37 -,54* -.09 -.10 -.12 .17 -.18 .35 -.40 ,53* .15 ,71** -.41 20 Maximum .25 .12 .06 .45 -,73** .16 -,67* -.10 -.41 .29 -.52 -.25 .31 -.36 .48 .23 ,84** ,65* .36 Mean Te am Performance
21 Quality .21 .26 .16 .39 .20 .19 .00 ,55* ,64* ,58* .01 ,78** .06 -.02 -.04 .02 -.15 -.16 -.16 -.22 22 Quantity .22 .35 .07 .41 .01 .26 -.10 ,52* ,61* .49 .08 ,72** -.06 -.15 -.08 -.15 -.05 -.17 -.09 -.11 ,86**
23 Communication .12 .35 -.07 .32 .02 .13 -.11 .43 .34 .37 -.04 .45 .14 -.22 .16 .06 .12 -.18 .04 .01 ,73** ,89**
24 Overall .16 .35 .02 .38 -.15 .33 -.29 .47 ,62* .51 .13 ,77** .13 -.24 .14 .00 .13 -.27 .16 -.04 ,88** ,93** ,86**
Te am Siz e .09 .27 -.14 .32 -.03 -.01 -.20 .37 -.18 .23 -.46 -.15 .50 -.11 .52 .42 .46 .27 .06 .49 .25 .28 .46 .22
*p<.05, **p<.01 (one-tailed)
The next step we took to gather results was to calculate correlations between all five of the individual personality variables, team performance, and communication scores, with controlling for team size (see Table 2).
Hypothesis 1
The results supported our first hypothesis, our results showed that mean conscientiousness is in fact significantly correlated with the overall team performance rating at p<.05, r=.68. This can be seen in Table 2, where we linked the correlations between individual personality variables with team performance. In this table we also controlled for team size. In both Table 2 and Table 1 we show values for the correlations between conscientiousness and performance, you will notice that these values differ in each table, yet they are both statistically significant measures. This slight difference is caused by the controlling for team size which was done in Table 2 and not in Table 1. The results here show that the average level of conscientiousness in a team has a positive relationship on the teams’ performance ability.
From our sample, none of the other mean measures (extraversion, agreeableness, emotional stability, and openness to experience) reported significant correlations with the overall performance rating.
Hypothesis 2
Our second hypothesis, for which we measured the variance of the individual personality scores, was not supported by our sample. We had predicted that a group having lower scores of variance for the conscientiousness scale, thus a more homogeneous group, would have higher scores of performance. This hypothesis would have been supported by a negative correlation between conscientiousness and performance. This thought was derived from the work of van Vianen & De Dreu (2001) who suggested that teams having heterogeneous conscientiousness would face more problems getting team tasks accomplished, which implies the likelihood of lower performance ratings. While our measures of emotional stability and autonomy did show negative correlations, neither of the measures was statistically significant. Our results did not support our hypothesis. From our sample we cannot conclude that teams having members which are more similar will perform better than teams which have less similar members.
TABLE 2
Correlation among Each of the Big Five Personality Traits with Controlling for Team Size (N=11)
Extraversion Team Performance Communication
Mean .15 .09
Variance .31 .26
Minimum .05 -.01
Maximum .33 .21
Agreeableness
Mean -.15 .04
Variance .34 .15
Minimum -.26 -.02
Maximum .43 .31
Conscientiousness
Mean .68* .48
Variance .48 .30
Minimum .26 .22
Maximum .83** .6*
Emotional Stability
Mean .02 -.13
Variance -.23 -.19
Minimum .04 -.11
Maximum -.11 -.17
Autonomy
Mean .04 -.12
Variance -.35 -.36
Minimum .15 .01
Maximum -.17 -.28
Communication .87**
*p<.05, **p<.01 (one tailed)
We analyzed the relationships between the different personality variables and performance, with specific focus on the conscientiousness scale. The last analysis we did was concerning the relationship between conscientiousness and team performance where we tested to see if communication has a mediating effect on team performance.
Hypothesis 3
In order to test an item for potential mediating effects, there must first be a significant relationship between the independent variables and the criterion (Baron & Kenny 1986).
In our study there was only one measure of conscientiousness which had a significant relationship with the communication measure. This was between the score of
conscientiousness-maximum and communication; we have chosen based upon theory from Baron & Kenny (1986) to look for the mediating effects that communication may have on performance and conscientiousness. In Table 3, Model 1 is a regression when communication is the dependent variable and conscientiousness is the independent variable. This regression analysis shows us that there is a relationship between communication and conscientiousness, but does not inform us which factor influences which. In Model 2 a regression pin points and examines the effect that conscientiousness has on performance. Here conscientiousness is again the independent variable and performance is dependent. This measure showed a significant relationship between the maximum score for conscientiousness and performance.
Then Model 3 was calculated with conscientiousness as the independent variable, performance as the dependent variable, and communication as the mediator. This shows whether the relationship between conscientiousness and performance is direct or indirect in nature. As can be seen in Model 3, when communication was added to the model, the measure of conscientiousness remains significant.
By adding communication to the model we can test its influence on the relationship between conscientiousness and performance. The results show that communication does act as a mediator between conscientiousness and performance; thus, the independent variable had an effect on the dependent variable via the mediator. Conscientiousness has a direct effect on performance. And conscientiousness has an indirect effect on performance via mediation by communication. When communication is a mediator the relationship between conscientiousness and performance becomes stronger. Because the direct relationship between conscientiousness and performance remains significant after communication is added to the equation shows us that it is an indirect mediator. Baron & Kenny (1986) explain in their study that perfect (direct) mediation is present when the independent variable no longer has any effect once the mediator is controlled.
While our study does not show a perfect mediating effect, Hypothesis 3 of our study is supported.
TABLE 3 Mediating Effects of Communication on the Conscientiousness-Performance Relationship
Model 1 B ∆ R2 Total R2
Team Size 0.22 0.22
Conscientiousness Max 1.07 0.28* 0.50*
Model 2
Team Size 0.05 0.05
Conscientiousness Max 1.02*** 0.66*** 0.7***
Model 3
Team Size 0.05 0.05
Communication 0.41*** 0.73**** 0.78****
Conscientiousness Max 0.59** 0.14** 0.92****
b= unstandardized regression coefficients
****p<.001, ***p<.01, **p<.05, *p>.1
Of the three hypotheses tested in this study, Hypothesis 1, and 3 were supported.
Discussion
The meta-analysis by Barrick & Mount (1991) which measured the relation of the Big Five to job performance criteria for five occupational groups, showed conscientiousness as being a
more restricted role.We decided to focus on conscientiousness because of the support it has had in this and other studies; thus leading us to the development of the three tested hypotheses.
It is worth noting that due to the great amount of time our sample teams spend working together we could make a good argument in favor of using any of Steiner’s (1972) types of tasks (see Unitary Tasks). Yet to manage the scope of this research we have chosen to limit our measurements to those outlined in our section on team tasks. Had we chosen to place emphasis on disjunctive and conjunctive tasks we would have needed to utilize the maximum and minimum scores to measure these factors. This is because these types of tasks see performance as determined by one member of the team, either the strongest or weakest individual. These measures are certainly of interest in such research. Yet our choice of which types of tasks to highlight is in accordance with a general principle taught to these student teams. It is worth noting, however, that since both additive and discretionary tasks are calculated using mean and variance respectively, this may cause the overall results of the research to lack diversity. This could create the impression of socially desirable responses, an issue addressed later in this section.
With hypothesis 1 we had expected mean levels of conscientiousness to have a significant correlation with team performance based upon the results of similar studies. The most significant finding in a study by Barrick & Mount (1991) found conscientiousness to consistently be a valid predictor for all of the occupational groups and all criterion types they studied in their 1991 meta-analysis of the Big-Five personality dimensions and job performance. Our hypothesis has been supported in our study, which is also in sync with the results of the study done by Barrick & Mount (1991).
On the other hand our Hypothesis 2 was not supported. We anticipated that groups with similar levels of conscientiousness would perform better than groups with less homogenous conscientiousness levels; our results did not prove this to be true. It is unclear from our results why the logical statements from van Vianen & De Dreu (2001) explaining that a more homogenous level of conscientiousness in a group will create less disagreement and therefore higher performance when working together were not supported by our study. To clarify this, further research could be aimed to test if well-rounded groups have higher performance levels than homogeneous groups. It may not be necessary for all team members to have an equal, or homogenous, level of conscientiousness; but rather that there be some sufficient level necessary to get the tasks accomplished.
Hypothesis 3 was supported by our data. We expected that communication would have a mediating effect on the relationship between conscientiousness and performance. Our results show communication having a positive, albeit mild, influence on this relationship. We used several variables to measure conscientiousness; we chose to use conscientiousness-max in our study since it was the only measure that had positive regression results with communication. The results of Table 3 may imply a direct relationship between communication and performance independent of conscientiousness. Barrick & Mount, in their study, express their confidence that this is a fruitful area for additional research. They also point out that it is an important issue to understand whether the effects of conscientiousness on performance are direct or indirect, or both. (1991).
We would like to acknowledge the weaker points in this study and with that seek to provide advice for future studies. Initially, the accuracy of our results may be influenced by the size
of our sample. We had some shrinkage of our sample; there were a number of unusable respondents since not all members of all teams chose to participate. Since it is difficult to find samples with a large number of teams this study considered hypotheses and methods used in several other studies conducted on team personality (Barrick et al. 1998, van Vianen and De Dreu, 2001). We feel that while our sample was limited, we have tapped a very useful resource by studying student teams. This is a resource which should be considered for future studies. Currently there is a trend towards team/group work in the classroom as well as in the workplace. Student teams should not be overlooked; to gain a larger sample one could increase the time frame of the study as to include students over the course of several years.
Another factor which we should not overlook in this assessment of our study is the possibility of respondents reporting socially desirable answers. Our survey was based upon a scale from 1-5, 1 representing strongly disagree, and 5 being strongly agree. When we look at the overall results of our survey there is an overwhelming majority of responses 3 and higher. Ones &
Viswesvaran (1998) show in their study that: “Their data from real-world job applicants confirms that criticizing personality scales because of potential response distortion by applicants is making much ado about nothing” (p. 266). In relation to our study, while it is plausible that respondents may have given some socially desirable answers to the Organizational Big Five Inventory, this is not sufficient to diminish the worth of this study.
Particularly considering Ones & Viswesvaran’s (1998) explaination that people seem convinced that this is a problem without any evidence of problems of social desirable answers effecting criterion-related validity of personality measures.
In addition to this, our results for zero-order correlations do not follow the same trend as the results from van Vianen & De Dreu (2001) or Barrick et al. (1998). In both of their reports their correlations become smaller with each different team level correlation measure. Barrick et al. reported .40, .22, and .16. Van Vianen & De Dreu had two separate samples in their study; the scores for their student team sample were .48, .37, and .22. For the drilling team sample the scores were .42, .30, and .19. The results of their studies suggest that different team level traits were operationalized differently (van Vianen & De Dreu, 2001). Looking at our results .50, .26, .26, the scores didn’t decrease from the second measure: same methods- different traits (i.e. all mean scales correlated with the mean measures from other scales) to the third measure: different traits-different methods (i.e. mean for extraversion correlated with variance, minimum, and maximum for the other scales). This implicates that, unlike in van Vianen & De Dreu and Barrick et al, it does not matter if we based our measurements on within trait or different trait because they score evenly. This may be a result of our small sample size.
In closing, our research has reinforced the notion that it is important to have knowledge about traits of teams and team composition because of the statistically significant results which have demonstrated a positive relationship between team personality, most specifically conscientiousness, and performance. We feel that this is an area of research which is important for the success of the growing number of organizations which rely more now than ever upon the use of teams. The abundance of self-managed teams in the work place creates new questions about the psychology of group performance (Barry & Stewart 1997). It seems that we have identified a useful renewable sample group; we have shown that students learning in self-managed teams can be a valuable resource for future studies. We hope to have encouraged further research on the issue of team performance and personality; we look forward to future studies which will provide the field with increased insight into the successes and failures of teams.
Reference List
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), 1173-1182.
Barrick, M.R., Mount, M.K. (1991). The Big Five Personality Dimensions and Job Performance: a Meta-Analysis. Personnel Psychology, 44, 1-26.
Barrick, M.R., Stewart, G.I., Neubert, M.J., & Mount, M.K. (1998). Relating Member Ability and Personality to Work-Team Processes and Team Effectiveness. Journal of Applied Psychology, 83. (3), 377-391.
Barry, B. & Stewart, G. L. (1997). Composition, Process, and Performance in Self-Managed Groups: The Role of Personality. Journal of Applied Psychology, 82, 62-78.
Deeprose, D. (1995). The Team Coach: Vital New Skills for Supervisors and Managers in a Team Environment. NY, USA: AMACOM.
De Raad, B. & Perugini, M. (2002). Introduction. In B. De Raad & M. Perugini (Eds.), Big Five Assessment. Göttingen, Germany: Hogrefe & Huber Publishers.
Glassman, S. (2000). Role Play. Sports Illustrated for Women, 2 (4).
Hendriks, A., Hofstee, W., De Raad, B. (2002). The Five-Factor Personality Inventory:
Assessing the Big Five by means of brief and concrete statements. In B. De Raad &
M. Perugini (Eds.), Big Five Assessment (79-108). Göttingen, Germany: Hogrefe &
Huber Publishers.
Hofstede, G. (1984). Culture’s Consequences International Differences in Work-Related Values (Abridged Addition). CA, USA: SAGE Publications Inc.
Hogan, R.T. (1991). Personality and personality measurement. In M.D. Dunnette & L. M.
Hough (Eds.), Handbook of industrial and organizational psychology, Vol. 2: pp. 873- 919.
Hough, L.M., Eaton, N.K., Dunnette, M.D., Kamp, J.D. & McCloy, R.A. (1990). Criterion- Related Validities of Personality Constructs and the Effect of Response Distortion on Those Validities. Journal of Applied Psychology, 75. (5), 581-595.
Huseman, R. C., Hatfield, J. D., Boulton, W. R., Gatewood, R. D. (1980). Development of a Conceptual Framework for Analyzing the Communication-Performance Relationship.
Academy of Management Proceedings, 178-182.
Mohammed, S., Mathieu, J.E., Bertlett, A.L. (2002). Technical-administrative task
performance, leadership task performance, and contextual performance: Considering the influence of team- and task-related composition variables. Journal of Organizational Behavior, 23. (7), 795-805.
Mount, M.K., Barrick, M.R., Stewart, G.L. (1998). Five-Factor Model of Personality and Performance in Jobs Involving Interpersonal Interactions. In Hogan, J. (Ed.) Human Performance Personality & Job Performance, 11. (2/3), 145-165. London, England:
Lawrence Erlbaum Associates.
Mount, M.K., Barrick, M.R., Strauss, J.P. (1999). The Joint Relationship of
conscientiousness and Ability with Performance: Test of the Interaction Hypothesis.
Journal of Management, 25. (5), 707-721.
Ones, D.S. & Viswesvaran, C. (1998). The Effects of Social Desirablility and Faking on Personality and Integrity Assessment for Personnel Selection. Human Performance, 11. (2/3), 245-269.
Steiner, I.D. (1972). Group Processes and Productivity. NY, NY: Academic Press, Inc.
Van Oudenhoven, J.P., Prins, K.S., Bakker, W., Schipper, M., Tromp, N. (2003).
Organisatiecultuur als persoonlijkheid: een op de Big Five gebaseerde vragenlijst, Gedrag & Organisatie, 16, 355-369.
Van Vianen, A.E.M., De Dreu, C.K.W. (2001). Personality in Teams: Its Relation to Social Cohesion, Task Cohesion, and Team Performance. European Journal of Work and Organizational Psychology, 10. (2), 97-120.
Wanous, J.P., Reichers, A.E. & Hudy, M.J. (1997). Overall Job Satisfaction: How Good Are Single-Item Measures? Journal of Applied Psychology, 82. (1), 247-252.
Witt, L.A., Burke, L.A., Barrick, M.R., Mount, M.K. (2002). The Interactive Effects of
Conscientiousness and Agreeableness on Job Performance. Journal of Applied Psychology, 87. (1), 164-169.
Witt, L. A. and Ferris, G. R. (2003). Social Skill as Moderator of the Conscientiousness- Performance Relationship: Convergent Results across Four Studies. Journal of Applied Psychology, 88. (5), 809-820.
