Inclusion in diverse work teams : the moderating role of climate for inclusion from an optimal distinctiveness perspective

Inclusion in Diverse Work Teams: The

Moderating role of Climate for Inclusion

from an Optimal Distinctiveness Perspective

ROBIN HARDERWIJK

ABSTRACT –

This study aims to increase understanding of diversity of gender, tenure, and age and its

relationship with collective efficacy, team performance, and team innovation. Elaborating on Brewer’s

Optimal Distinctiveness Theory (1991) and Shore and colleagues’ notion of inclusion in diverse

contexts (2011). Results show that climate for inclusion affects the fulfilment of the need for

belongingness to help fulfilment of the need for uniqueness. With these needs balanced, it appears to

be possible to foster self-efficacy-feelings of team-members. In order to present results, one scale was

developed and two studies were conducted on 32 supervisor rated teams and a group of 95

participants.

University of Amsterdam

MSc Business Studies

Faculty of Economics and Business

Course:

6314M0253 – Master’s Thesis Leadership and Management

Semester 2, 2014-2015, Period 6

Coördinator:

Dr. Karin Venetis

Supervisor and first reader:

Dr. Claudia Buengeler

Second reader:

Dr. Wendelien van Eerde

Due Date:

June 29, 2015

Date:

June 29, 2015

Student number:

1066680

Word Count:

12981

(ex. Appendice)

Contents

Statement of originality ... 3

Introduction ... 4

A Model of Team Diversity, Work Team Performance and Optimal Distinctiveness ... 5

Optimal Distinctiveness ... 7

Conceptualizing two models of the influence of CfI on need fulfilment in diverse work teams ... 7

Overview of studies ... 10

Method ... 11

Procedure and Sample Study 1 ... 11

Procedure and Sample Study 2 ... 11

Measures ... 12

Measures study 1 ... 12

Scale development optimal distinctiveness in fulfilment of the NfU and the NfB ... 12

Performance and CfI ... 13

Measures study 2 ... 14

Diversity ... 14

Fulfilment of NfU and NfB ... 14

Performance and CfI ... 15

Control variables ... 15

Data Aggregation justification ... 15

Results ... 18

Results Study 1 ... 18

Results Study 2 ... 19

Discussion ... 27

Theoretical Implications ... 27

Practical Implications ... 29

Limitations and Future Directions ... 31

General Conclusion ... 32

References ... 33

Appendices ... 39

Appendix 1 ... 39

Appendix 2 ... 68

Appendix 3 ... 99

Appendix 4 ... 114

Appendix 5 ... 124

Appendix 6 ... 129

Appendix 7 ... 137

Words of thanks

The author would like to thank all involved in the forthcoming of this research and thesis. Special

thanks goes out to Claudia Buengeler for her supervision, helpful insights and inexhaustible support

and patience. A warm thanks goes out to Peter and Cathy Harderwijk, both for doing quite a successful

job in being the most loving and caring parents I know, and of course for being the main sponsors of

my studies. Also thanks goes out to everyone who was willing to participate in and/or give helpful

insights on everything that has been written for this project. I would like to thank JC Manticore, The

Collegium Illustrissimum Hermandad and of course everyone else concerned with Unitas S.R., older

friends and other groups and gatherings I have been involved with, during or before my studies. You,

and all other friends, have made my time as a student the best time I have had. Thank you Marjan for

your love, care, support and advice.

Statement of originality

This document is written by Robin Harderwijk who declares to take full responsibility for the

contents of this document. I declare that the text and the work presented in this document is original

and that no sources other than those mentioned in the text and its references have been used in

creating it. The Faculty of Economics and Business is responsible solely for the supervision of

completion of the work, not for the contents.

Introduction

“After more than two decades of diversity research, four decades of antidiscrimination legislation, and

extraordinary media attention to diversity, discrimination and exclusion in organizations persist” (Bell,

2007, p. 3). Discrimination and exclusion of minority social groups can have negative consequences

for employee health such as depression and burn-out (Williams, Neighbors, & Jackson, 2003). When

employees feel excluded from their teams they are at risk for lower productivity, absenteeism and

higher turn-over intentions (Cox, 1994) which is likely to negatively influence organizational outcomes

(Chen & Velsor, 1996) and performance (Jehn & Bezrukova, 2004).

Although the past focus on “problems” associated with diversity, as described above, has

generated many meaningful and informative theories and empirical studies, and continues to do so

(Shore et al., 2011), the field has since then evolved tremendously. Scholars have increasingly focused

on more positive outcomes of diversity on team and organizational performance and work processes

(Gonzales & DeNisi, 2009 in; Shore et al., 2011, Knippenberg, De Drue & Homan, 2004).

Already much earlier, Cox (1991), stressed the importance of the search of ways to integrate

diverse individuals in organizations (Thomas & Ely, 1996). Within an ever more globalizing society it

appears to be an important issue that the problems of diversity have to be overcome (Jehn &

Bezrukova, 2004) and that the benefits of diversity are used for increasing performance (Ely, 2004). In

order to sustain their competitive advantages in a global market, companies are increasingly

recognizing the need to leverage their employees’ diversity (Jehn & Bezrukova, 2004; Offerman &

Gowing, 1990; Thomas & Ely, 1996; Yaprak, 2002), especially within the international trend toward

increased immigration, expatriatism, the globalization of firms (Jehn & Bezrukova, 2004; Johnson,

2002; Yaprak, 2002) and the North American and European trend towards an aging workforce and a

greater representation of woman and minorities in the workplace (Friedman & DiTomaso, 1996;

Gorski, 2002; Jackson, Stone, & Alvarez, 1993; Jehn & Bezrukova, 2004).

“Diversity appears to be a double-edged sword, increasing the opportunity for creativity as

well as the likelihood that group members will be dissatisfied and fail to identify with the group”

(Milliken & Martins, 1996, p. 403). Nonetheless company leaders often presume that greater diversity

will automatically lead to (often unspecified) benefits while ignoring the complicated issues of

managing work teams with high diversity (Jehn & Buzrakova, 2004; Kersten, 2000; Shaw &

Barret-Power, 1998). Thus, further understanding of how diversity can be used to achieve greater

performance is of high importance (Shore et al., 2011).

A Model of Team Diversity, Work Team Performance and Optimal Distinctiveness

Past results of diversity research could be considered contradictory. In particular, because

some prior studies show positive effects (e.g. Hoffman & Maier, 1961; Jehn, Bezrukova, 2004; Jehn et

al., 1999; Watson, Kumar, & Michaelson, 1993), whereas others show that diversity has negative

effects (Bell, 2007) or is unrelated to performance (e.g. Ancona & Caldwell 1992; Bell, 2007; O’Reilly

& Flatt, 1989; Steiner, 1972). An explanation of this apparent contradiction can be found in the

complicatedness of the used research methods, due to the interactional effects, diversity appears to

have with a variety of other team and organizational factors (William & O’Reilly, 1998; Jehn &

Bezrukova, 2004). Even though a few studies have looked carefully at the effects of these factors on

the relationship between diversity and outcomes (cf. Chatman, Polzer, Barsade, & Neale, 1998; Jehn

& Bezrukova, 2004; Jehn et al., 1999). Despite widespread public opinion in favor of the virtues of

diversity (Ely, 2004), a consistent positive link between diversity and work team, business, or

organizational performance has not been established (Ely, 2004; Jehn & Bezrukova, 2004; Shore et

al., 2011). Thus, consistent with calls for research on the moderating factors that help explain this

inconsistent relationship (Cho & Mor Barak, 2008), this study will explore under which circumstances

an individual worker and/or work team will be able to optimally use diversity for heightened

performance.

In the diversity literature there is support for the advantages of experiencing fulfilment of

both the needs for belongingness and uniqueness. As an example, Friedman, Kane and Cornfield

(1998) showed that minority members (who experience a fulfilment of their need for uniqueness) with

developed networks (which makes them experience fulfilment of their need for belongingness) report

a high level of career optimism. Hitlan, Clifton and DeSoto (2006) found that workplace exclusion can

be very harmful. Especially the rejection by supervisors or co-workers is detrimental to work attitudes

and psychological health. When both the need for uniqueness and the need for belongingness remain

unfulfilled, one might not feel less included, but even excluded. This is when an individual is not

treated as an organizational insider with value to the outcomes of a work team (Shore, et al., 2011). A

thwarted NfB can have serious harmful cognitive, emotional, behavioral and health outcomes (e.g.,

Baumeister, DeWall, Ciarocco, & Twenge, 2005; Blackhart, Nelson, Knowles, & Baumeister, 2009).

Exclusion research has focused on the need for belongingness through social rejection (Hitlan, Clifton,

& DeSoto, 2006). But when the need for uniqueness is unmet due to low recognition (for example,

when colleagues do not value characteristics such as specific knowledge or unique skills), this should

also contribute to feelings of exclusion (Shore, et al., 2011).

Therefore, the question to be answered in this research is whether the fulfilment of both the

need for uniqueness and the need for belongingness will lead to inclusion in diverse (compared to

homogeneous) teams, and thus have positive effects on employee and team outcomes, such as

performance, as argued by Shore and colleagues (2011).

Using Williams and O’Reilly (1998) as a guide to the many relevant aspects on which

individuals can differ, diversity is broadly defined as “any attribute that people use to tell themselves

that another person is different” (p81). Traditionally, diversity is conceptualized as easily visible

differences such as age or gender (Hicks-Clarke & Illes, 2000; Jehn & Bezrukova, 2004). Nonetheless

individuals may also differ on less visible characteristics, such as tenure with the organization. In this

study, team diversity is considered along three demographic dimensions. Gender and the aspects of

age and (team-)tenure, since according to prior studies these can be considered adequate and

conclusive measures of diversity, and thus are able to represent a larger spectrum of diversity aspects

(cf. Acquavita, Pittman, Gibbons, Castellanos-Brown, 2009; Cho & Mor Barak, 2008; Ely, 2004; Jehn &

Bezrukova, 2004; Nishii, 2013; Roberson, 2006; Shore et al., 2011; Shore et al., 2009).

Shore and colleagues (2011) proposed that, when managed correctly, diversity might ensure

potential value to organizations. Particularly in areas such as increased creativity (McLeod, Lobel &

Cox, 1996) and decision-making (Maznevski, 1994). It appears that especially when members of work

teams feel included in the social environment of the team, more positive effects of diversity show

(Roberson, 2006). It is suggested that a way of influencing these feelings and attributions of inclusion

is through the fulfilment of the Need for Belongingness (NfB) and the Need for Uniqueness (NfU)

(Shore et al., 2009). By doing so they built upon the model of Optimal Distinctiveness as proposed by

Brewer (1991). This theory (Brewer, 1991, p477) explains the tensions associated with “human needs

for validation and similarity to others (on the one hand) and a countervailing NfU and individuation

(on the other).” It is clear that, contrary to the negative organizational and employee outcomes

coming from exclusion and discrimination in diverse social groups, inclusion could lead to positive

organizational and employee outcomes (Chen & Velsor, 1996). Nonetheless, limited research has been

conducted on how inclusion can be reached (Cox, 1994; Chen & Velsor, 1996; Jackson & Alvarez, 1992;

Shore et al., 2011). There are some strong implications that Optimal Distinctiveness relies on the

balance of the fulfilment of these two needs (e.g. Hornsey & Hogg, 1999; Hiel & Mervielde, 2002;

Leonardelli, Pickett, & Brewer, 2010). And thus, the conceptualization of inclusion promises to

advance the diversity literature on several fronts (Shore et al., 2011). Despite of the frequent

assumption that non-similar individuals are likely to be viewed unfavourably (Abrams & Hogg, 2002),

the inclusion framework posits that, within teams, members can be valued for unique characteristics

and feel included and belonging to the team at the same time (Shore et al., 2011).

Optimal Distinctiveness

There is a clear need for further insight on the influence of diversity on team outcomes and

climates under which favourable outcomes (e.g. performance, efficacy, innovation) could arise (Shore,

et al., 2011). Also when it comes to Brewers (2011, p. 90) words: “If social differentiation and

intergroup boundaries are functional for social cooperation, and social cooperation is essential for

human survival, then there should be psychological mechanisms at the individual level that motivate

and sustain in-group identification and differentiation.” The Optimal Distinctiveness Theory (ODT)

describes the two apparent opposing human needs that, when fulfilled, lead to perceptions of

inclusion (Brewer, 1991). The NfB and the NfU govern the relationship between the membership in

social groups and the self-concept. She proposes that “the first is a need for assimilation and inclusion,

a desire for belonging that motivates immersion in social groups” (Brewer, 2011, p. 90) and the second

is a “need for differentiation from others that operates in opposition to the need for immersion”

(Brewer, 2011, p90). When the NfB is satisfied, the NfU grows (Brewer, 1991), and when both needs

are high, team members will feel included (Shore, et al., 2011). Whereas this inclusion is defined as

“the degree to which an employee perceives that he or she is an esteemed member of the work team

through experiencing treatment that satisfies his or her needs for belongingness and uniqueness”

(Shore, et al., 2011, p. 1265). Building on ODT (Brewer, 1991), this definition is distinct from others in

a way that it explicitly states that both needs need to be satisfied in order to foster inclusion.

Conceptualizing two models of the influence of CfI on need fulfilment in diverse work teams

As has been previously described, it is important to test the conceptual model of Shore and

colleagues (2011) whether individuals who see both their NfB and NfU fulfilled, do perform better

than those who do not find fulfilment of above needs.

Hypothesis 1: Performance in diverse (compared to homogeneous) teams is mediated by

fulfilment of the NfU and NfB, whereas lower levels of need fulfilment will lead to lower

performance in diverse work teams.

When looking at the existing literature on inclusion in diverse work teams two general themes

are apparent when it comes to definitions of inclusion in reference to Optimal Distinctiveness Theory.

For example: “when individuals feel a sense of belonging, and inclusive behaviors such as eliciting and

valuing contributions from all employees are part of the daily life in the organization” (Lirio, Lee,

Williams, Haugen, & Kossek (2008, p433) or Holvino, Ferdman and Merrill-Sands (2004, p249) who

define an inclusive organization as “one in which the diversity of knowledge and perspectives that

members of different groups bring to the organization has shaped its strategy, its work, its

management and operating systems, and its core values and norms for success.” Or “the extent to

which employees believe their organizations engage in efforts to involve all employees in the mission

and operation of the organization with respect to their individual talents” (Avery, McKay, Wilson, &

Valpone, 2008, p. 6).

Reappearing themes in these definitions are “sense of belonging”, “insider” and “accepted”.

Three terms that were combined in Nishii’s (2014) formulation of Climate for Inclusion (CfI). Because

due to the pervasive awareness of the need for proper management of the concerns and possible

benefits related to diversity, organisations have developed different approaches to manage diversity

in work teams (Roberson, 2006). Diverse work teams occupy a pivotal role in organizational outcomes

(Jackson & Ruderman, 1995) and are commonly defined as “interdependent collections of individuals

who share responsibility for specific outcomes for their organizations” (Sundstrom, Meuse, & Futrell,

1990, p120). And thus, the impact of diversity on performance through optimal distinctiveness will

only be reached when both the NfU and the NfB will be optimally fulfilled. In current conviction it is

clear that to really manage both adverse as beneficial effects of diversity, organizations have to create

environments that are inclusive to all workers (Ferdman & Davidson, 2004; Shore, et al., 2011; Nishii,

2013). Although promising evidence for the beneficial effects of inclusive environments exists, the

body of empirical testaments can still be considered scarce (Nishii, 2013). Nonetheless, Nishii (2013,

p. 1767) made clear that the “important aspect of inclusive climates is that they facilitate the

engagement of whole selves”, which includes the expression of multiple identities (Rothbard &

Ramarajan, 2009; Nishii, 2013). Thus it increases the probability that cross-cutting ties emerge within

teams with high diversity, which could lead to more efficacy and more performance (Vignoles, Regalia,

Manzi, Golledge & Scabini, 2006).

This means that higher levels of diversity serve as a predictor for team members to find a

balance in the fulfilment of their needs for uniqueness and belongingness, as long as this relationship

is moderated by higher levels of the CfI, as can also be seen in Figure 1. Which brings us to the following

hypothesis:

Hypothesis 2: A diverse (compared to homogeneous) work team’s CfI moderates the strength

of the relationship between team diversity and need fulfilment in such a way that there will

be more need fulfilment if there is a stronger CfI.

However, research with a focus on (dis)similarity of team members within work teams has often been

mixed (Shore et al., 2011), which suggests that demographic (dis)similarity may not always promote a

sense of belongingness on its own (cf. Riordan & Wayne, 2008). Nonetheless it can be deductively

reasoned that when in a highly diverse team, one’s uniqueness is more salient and therefore the NfU

will be fulfilled. Nonetheless, a singular focus on fulfilment of the NfU, or differentiation, can lead to

interpersonal interactions involving the overreliance on stereotypes and even segregation from the

work team (Ely & Thomas, 2001). Thus both the fulfilment of NfU and NfB should be high. Only then,

the beneficial outcomes as proposed by Shore, and colleagues (2011) will arise, as can also be seen in

figure 2. Which leads us to the following two hypotheses:

Hypothesis 3: Within a diverse (compared to homogeneous) work team, the fulfilment of the

NfU will mediate the relationship between the diversity of a team and its performance.

Hypothesis 4: A diverse (compared to homogeneous) work team’s fulfilment of the NfB will

moderate the relationship between the fulfilment of the NfU and the team’s performance in

such a way that by making work team members feel unique and valued will make them

perform.

Nonetheless, in line with research that shows that creativity is enhanced through diverse groups that

engage in collaboration (Levine & Moreland, 2004) and in groups where heterogeneous views are

shared (Simonton, 2003), it should be considered how providing conditions to have the NfB fulfilled in

order to support the relationship between fulfilment of NfU and performance outcomes. Or as Nishii

Diversity

Need fulfilment

Uniqueness & Belongingness

Climate for

Inclusion

Team

Performance

Figure 1. The moderating effect of CfI on the mediated relation between Diversity, need

fulfillment, and performance.

(2010 in; Shore, et al, 2011) showed, CfI leads to interpersonal integration of diverse employees and

thus fulfilling their needs for belongingness. This leads to the next hypothesis:

Hypothesis 5: The relation between diversity and team performance will be mediated by

fulfilment of the NfU, after team members experienced fulfilment of the NfB and CfI in such a

way that higher CfI will lead to a stronger moderating effect of NfB on the relationship

between fulfilment of NfU and team performance.

Overview of studies

The described hypotheses were researched in two different research setups. First, it was of

interest whether people would expect the fulfilment of both the NfU and the NfB to be of influence

on their performance, especially when the moderating effect of CfI would be taken into account. In

study 1, a convenience sample of mainly university students and starters on the job market read a

scenario about working in a generally homogenous or diverse team with either a stronger or weaker

CfI. This generates an experimental 2 (diversity: high vs. low) X 2 (CfI: high vs. low) experimental

between-subjects design.

In study 2, it was of interest on whether performance (as measured by supervisor performance

ratings) was influenced by team diversity, how fulfilment of the NfU would mediate this relationship,

Diversity

Need fulfilment

Uniqueness

Need fulfilment

Belongingness

Team

Performance

Climate for

Inclusion

Figure 2. The positive relationship between diversity and team performance is mediated by

fulfilment of the NfU when team members also experience fulfilment of the NfB and CfI in such a

way that higher CfI will lead to a stronger moderating effect of NfB on the relationship between

fulfilment of NfU and team performance.

and how the fulfilment of the NfB and CfI moderate the relationship between fulfilment of the NfU

and performance.

Method

Procedure and Sample Study 1

In this study, 97 participants participated (52 male (mean age = 34.245, SD = 15.224) and 43

femal (mean age = 29.953, SD = 11.737), total mean age = 32.323, SD = 13.870) which were

convenience sampled and randomly assigned to one of the four experimental conditions. In this

experiment informed consent was obtained from all participants. If the survey was completely filled

in, all participants had the chance to win one out of three €20 gift voucher for a popular web-shop.

The questionnaire could be filled in through Qualtrics software (Copyright © 2014 Qualtrics,

Provo, UT) or paper-based. Filling in the questionnaire took about 15 minutes. Based on language

preferences of the participants the questionnaires (to be found in Appendices 1 and 2) could be filled

in in both English and Dutch. Although all measures were originally in English, the original versions

were translated into Dutch by the author. Although all measures were originally in English, the original

versions were translated into Dutch by the author. To ensure consistency Brislin’s (1970) back

translation method was used, in which someone (proficient in both English and Dutch), not familiar

with the original items translated the Dutch translations back to English. This way, the translated items

were found to be consistent with the original items.

The distribution of participants over the experimental conditions can be found table 1.

Table 1. Age distribution over experimental conditions

Experimental Condition

1

2

3

4

N

13

16

15

14

AGE

M

29.643

36.176

32.438

31.533

SD

13.698

16.607

14.109

14.471

*) Category 1: Low diversity - low CfI; Category 2: High diversity - low CfI; Category 3: Low diversity

- high CfI; Category 4: High diversity - high CfI

Procedure and Sample Study 2

In this study, 188 people participated (whereof 31 supervisors and 157 employees, of which the age

and gender distribution is shown in table 4), clustered in 32 teams (mean within team participation

rate: 92%). The convenience sample of participants worked in 14 Dutch and international companies

both in the private and public sector, whereof team members worked either voluntarily or paid. After

obtaining informed consent, participants filled in a survey, distributed online through Qualtrics

software (Copyright © 2014 Qualtrics, Provo, UT) or paper-based. Filling in the questionnaire took

around 15 minutes. Based on participant’s language preferences the survey was offered in either

English or Dutch (see appendix 3, 4, 5

and 6). To ensure consistency Brislin’s (1970) back translation

method was used.

Data was gathered on the team level to ensure the possibilities of aggregation to the team

level. The average team size was 6.5 team members (SD = 4.73). As is shown by the high standard

deviation, there was a wide range in team size from 2 to 25 members. The participating team

members’ average age was 32 (SD = 11.171). Gender distribution for team supervisors were 35.7%

female, whereas team members were more evenly divided, with 43.8% being female.

Measures

All surveys consisted of questions to measure the various constructs of team diversity, need

fulfilment and CfI. All questions were rated on a 5-point Likert scale which consists of 1=Strongly

Disagree, 2=Disagree, 3=Neither Agree nor Disagree, 4=Agree and 5=Strongly Agree, unless stated

otherwise in the following section. Some items were reversed stated, and thus had to be recoded.

Missing data was handled through hotdeck imputation (Myers, 2011). All constructs were analyzed on

their frequencies to see whether missing values were present and whether it was appropriate to use

hotdeck imputation. All constructs had less than 10% missing values, which makes using hotdeck

imputation appropriate. The imputation replaced missing values with ‘donor’ values that matched the

‘donee’.

Measures study 1

Scale development optimal distinctiveness in fulfilment of the NfU and the NfB

In order to test the first model, a new scale had to be developed in order to fully measure the

fulfilment of the NfB and the fulfilment of the NfU. Items were generated deductively (Hinkin, 1998)

from a review of existing literature on (psychological) need fulfillment (e.g. Shore, et al., in print;

Stamper & Masterson, 2002; Ryan & Deci, 2000) and job-person, team-person and,

organization-person fit questionnaires (Kristoff-Brown, Zimmerman, & Johnson, 2005). The advantages of

deductive scale development is that it helps to ensure content validity, as the items are written to

reflect an existing construct definition. Nonetheless, in order to fully check content validity, several

(renowned) scholars on team diversity, CfI and optimal distinctiveness were asked to assess the

content validity of the items by matching them to the corresponding definitions of the two

dimensions. The retained items (rated on a 5 point Likert scale ranging from 1=Strongly Disagree,

2=Disagree, 3=Neither Agree nor Disagree, 4=Agree to 5=Strongly Agree), were administered to a

student sample of 203 respondents.

In terms of gender, 63.72% of the respondents were female, whereas in terms of age, the

mean age of participants was 22,21 (Min = 17, Max = 59, SD = 6.103). Participants team tenure, the

time participants were with their team, was on average 33.11 months (SD = 69.270). Their educational

level ranged from high-school (38%), vocational school (1.9%), Bachelor’s degree (35.2%), Master’s

degree (9.3%) to PhD (.9%). Of all participants, 63% worked part-time.

In line with the method of Field (2011), after conducting a scree plot of factor components,

three factors could be derived and by conducting exploratory factor analysis on the items suggested

that, after items with low (< .40) and double factor loadings had been dropped, a smaller set of 11

items loaded on the one factor with eigenvalues greater than .7. Internal consistency of the first set

of 42 items was α = .965.

One absolute goodness-of-fit test was assessed, which was the Root Mean Square Error of

Approximation (RMSEA). The

χ

_{goodness-of-fit indice is sensitive to sample size, thus it is}

recommended to use a relative goodness-of-fit measure (Bentler, 1990). Therefore the Tucker-Lewis

Index (TLI) and the Normed Fit Index (NFI) were used. Values smaller or around .05 are considered an

excellent fit for RMSEA and values higher or around .95 are indicative for an excellent fit for the

relative indices (Hoyle, 1995). Partial confirmatory factor analysis (PCFA) on the 42 items showed that

the one-factor structure – consisting of fulfilment of both the NfB and fulfilment of the NfU – exhibited

better fit (NFI = .985, CFI = .990, TLI = .970, RMSEA = .052) than a two-factor structure (NFI = .815, CFI

= .780, TLI = .743, RMSEA = .092). The one dimension containing four retained items, exhibited high

reliability (α = .847). The questionnaire used for the validation of the construct can be found in

appendix 7.

Performance and CfI

Performance was measured through the Self-Rated Job Performance scale (12 items; α = .911;

Judge, Erez, Bone & Thoreson, 2003). An example of these items is “When I try, I generally succeed”

(+). CfI is measured through the CfI scale (15 items; α = .828; Nishii, 2013). An example of these items

is “This team has a fair promotion process” (+). A correlation matrix of all variables can be found in

table 2, wherein it can be seen that there are significant correlations between fulfilment of both the

NfU and NfB (ODT) and performance, between CfI and ODT, and CfI and performance, which one

would expect according to the model of diversity and inclusion by Shore and colleagues (2011),

Table 2. Means, Standard Deviations, Correlations

Variables

M

SD

1

2

3

1

Diversity

0.520

0.502

-

2

ODT

3.380

0.921

0.047

-

3

Performance

3.447

0.703

0.092

0.613 *

-

4

CfI

0.530

0.501

-0.074

0.460 ** 0.246 *

**. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed).

N = 95

Measures study 2

Diversity

All demographic data used to calculate diversity indices were reports of the team members’

demographics gathered through team supervisors. The diversity measures of age and (team)tenure

were contained in two interval scales, whereas gender was contained in a nominal scale. The former

two were indexed using departmental standard deviations (SD); whereas SD was chosen over other

diversity measures since maximum SD does not increase with the size of a unit (Harrison & Klein,

2007). Due to the categorical nature of gender, however, SD is not an appropriate measure of

diversity. Thus, the Blau’s (1977) index was used in order to measure gender diversity. Blau’s index is

a commonly used index for diversity of categorical variables (Harrison & Klein, 2007). The index is

calculated through the formula 1 - Σpₐ

_{, where p is the proportion of unit members in the a}

_category.

The values of this index can range from zero to (a – 1)/a. Its maximum occurs when members of a

team, Vₐ’ are spread equally over all possible a categories (Harrison & Klein, 2007). This way of

conceptualizing gender diversity is, according to one of the latest meta-analyses, used in 23 out of 26

field studies concerned with gender diversity (Bell et al., 2011 in; Nishii, 2011). Implicit to the use of

Blau’s index is the assumption that the effects of gender diversity on performance are symmetric

(Harrison & Klein, 2007). This means that when, for example a group comprised of 40% women should

not be qualitatively different from a group with 60% women (Nishii, 2011). For these reasons, Blau’s

index was deemed to be an appropriate measure for gender diversity.

Fulfilment of NfU and NfB

The fulfillment of the NfB and NfU is measured through the Need fulfillment questionnaire

(10 items; α = .698; Shore, et al., in print). Exemplar items of the Shore, et al. (in print) questionnaire

are ”I belong in my team” (Belongingness; +) and “People in my team listen to me even when my views

are dissimilar” (Uniqueness; +).

Performance and CfI

Given that supervisors are considered “experts” with respect to evaluating the performance

of their subordinates, compared to self-rated performance (Viswesvaran, Ones, & Schmidt, 1996),

performance is measured through supervisor rated team performance (6 items; α = .926 ; Kirkman &

Rosen, 1991; exemplar: “This team meets or exceeds its goals”; +) and team innovation (4 items; α =

.695 ; Anderson & West, 1998; exemplar: “Team members often produce new services, methods or

procedures”; +). Team efficacy (3 items; α = .709; +) was measured through the Edmondson (1999)

scale with “Achieving this team's goals is well within our reach” as an exemplar item. Again, CfI is

measured through the CfI scale (15 items; α = .843; Nishii, 2013).

Control variables

As team size (measured within the supervisor questionnaire) can be considered an influencing

variable for team effectiveness, it has been controlled for (e.g. Jehn, 1995; Labianca, Brass & Gray,

1998; Somech, Desivilya & Lidogoster, 2009). Also task interdependence, due to its nature of having

(critical) influence on (team) performance (e.g. Campion, Medsker & Higgs, 1993; Cohen & Bailey,

1997; Stewart & Barrick, 2000), was controlled for (example: “I have a one-person job” (-); 5 items; α

= .796; Van der Vegt & Janssen, 2003) was used.

Data Aggregation justification

As the measured constructs were conceptualized at the team level, the mean r

values

were computed using an Interrater Agreement (IRA) and Interrater Reliability (IRR) estimates for

consensus composition constructs – tool, as prepared by Biemann, Cole and Voelpel (2012) in order

to detail the degree of agreement between the team members. Intraclass correlation coefficients

were also calculated (Bliese, 2000) to indicate the ratio of between-group and total variance (ICC

)

corrected for the average team size (Biemann, Cole, & Voelpel, 2012), the respective F-tests, and the

reliability of team members' average ratings (ICC

). Resulting in the values of .69 (r

), .23 (ICC

), and

.55 (ICC

) for CfI, .97 (r

), .24 (ICC

) and .57 (ICC

) for Belongingness and Uniqueness, and .83 (r

),

.01 (ICC

), and .04 (ICC

) for Efficacy. Since r

values were found to be above or close to the acceptable

point of .70 (Biemann et al., 2012), aggregation to the team-level was justified. For team performance

and innovation aggregation was not needed as these variables were already measured at the team

level.

Correlations

The means, standard deviations, and correlations of the variables of interest were calculated

(see Table 3), whereas a few significant correlations were found. In line with Shore and colleagues

(2011) framework of inclusion and diversity, CfI was positively correlated with the fulfilment of the

NfU, NfB and Efficacy. The fulfilment of the NfB was positively correlated with the fulfilment of the

NfU, Efficacy and Innovation. Efficacy was positively related with the fulfilment of the NfU. Also tenure

diversity was positively correlated with age diversity, which can be considered logically due to team

members with a higher age having had more time to work within a team contrary to team members

with a younger age.

Table 3. Means, Standard Deviations, Correlations

Variables

M

SD

1

2

3

4

5

6

7

8

1

Gender diversity

0.29

0.21

-

2

Age diversity

5.49

4.89

0.185

-

3

Tenure diversity

31.29

49.61

0.171

0.700 **

-

4

Team efficacy

3.80

0.35

0.042

-0.283

-0.126

-

5

Team performance

5.18

0.94

0.014

0.018

0.098

0.093

-

6

Team innovation

4.86

0.95

0.270

0.144

0.118

0.050

0.125

-

7

Fulfilment of NfU

3.99

0.36

0.146

0.842 **

0.171

0.466 **

0.125

0.322

-

8

Fulfilment of NfB

3.98

0.47

0.075

0.095

0.108

0.394

*

0.169

0.418 * 0.842 **

-

9

CfI

3.79

0.34

-0.046

0.004

0.086

0.438

*

-0.023

0.055

0.795 ** 0.756 **

**. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed).

N = 32

Results

Results Study 1

Hypothesis 1 predicted that high need fulfilment mediates the relationship between the

diversity in teams and team performance in such a way that diversity is positively linked to

performance through high levels of need fulfilment. Although the assumption that all constructs have

to correlate (Baron & Kenny, 1986) was violated, there was tested for mediation, and thus no

significant effects were found (M = 3.447; SD = .703), F(1, 123) = .2685, p = .605, η

_{= .002, one-tailed),}

nonetheless, a significant direct effect of need fulfilment on performance was found, F(2, 122) =

37.341, p < .05, η

_{= .380, one-tailed), which means Hypothesis 1 has to be rejected.}

The rejection of Hypothesis 1 makes it interesting to test Hypothesis 2. This hypothesis stated

that the mediated relationship of diversity and performance would be moderated by a team’s CfI. This

moderation would make that work teams with higher levels of CfI, experience a stronger relationship

between fulfilment of the NfU and the NfB and performance. CfI did explain a significant increase in

explained variance in the model, Δ η

_{= .265, F(2, 83) = 31.510, p < .05. Thus, CfI was a significant}

moderator of the relationship between the fulfilment of the NfU and the NfB and performance. The

gradient of the simple slopes is .610 (T = 2.366, P = .020), also see Figure 3. Although the relationship

between diversity and the fulfilment of needs is moderated by CfI (whereas high CfI has an even

stronger effect in circumstances with high diversity contrary to lower CfI), this is not indicating that

the direct relationship of diversity and performance disappears. This means that Hypothesis 2 can only

be partially confirmed.

Figure 3. Simple slopes of diversity predicting self-rated job performance for low and high CfI.

1

1.5

2

2.5

3

3.5

4

4.5

5

Low Diversity

High Diversity

Self

-rated

Jo

b

P

erfro

m

an

ce

Low CfI

High CfI

Results Study 2

The findings of the previous study suggested that, within experimental context, employees

who would experience higher CfI also experience higher fulfilment of their NfU and NfB, and that they

rate their performance even higher when those needs are fulfilled. Nonetheless it has not yet been

shown whether team performance actually increases under the predicted conditions. It is likely that

when team members are working in a highly diverse team, their NfU will be fulfilled, but only when

also the NfB is fulfilled teams will show higher performance and efficacy (i.e. moderated mediation;

Field, 2013). Furthermore this NfB is expected to be more fulfilled under the circumstances of a strong

CfI (i.e. mediated three-way interaction).

The main goal of this research was to examine how CfI influences the relationship of NfU and

NfB in diverse workteams. Also it is predicted that through CfI and need fulfilment, diverse teams will

have higher performance, innovation and efficacy. This effect is reached through the moderating

influence of CfI on the moderating influence of fulfilment of NfB on the mediation relationship of NfU

on team performance outcomes in a diverse context.

In hypothesis 3 it was predicted that the fulfilment of the NfU would mediate the relationship

between the diversity of a team and its performance. To test this, hierarchical linear regression

analyses were conducted. First, the control variables, Team Size and Task Interdependence, were

entered in the regression. Second, gender diversity (Blau’s index), age diversity and tenure diversity

(both standard deviation) were entered.

To analyse the relationship between the variables using a non-parametric resampling

approach, bootstrapping was used by using the process tool of Preacher and Hayes (2004). These

analyses were conducted through using the PROCESS tool by Andrew Hayes (2012), model 4, which

tests the indirect effect of the X on Y through Mj

= a

b

, the total indirect effect of X on Y through all

M = ∑

b

), and the direct effect of X on Y = c’

Whereas X is the three diversity measures, Y is three

performance outcomes and Mj is the proposed mediator NfU.

Nine analyses have been conducted to test for mediation of the three diversity measures (Age,

Tenure and Gender diversity) on the three performance outcomes (Efficacy, Team Performance and

Innovation). Although it appeared that not all variables were related (as they should be, according to

Baron and Kenny, 1986, and Field, 2013), there was tested for mediation. As can be seen in Table 4, it

appeared that only the relationship of the three diversity measures and efficacy were mediated by

fulfilment of the NfU at the .05 level of significance. Also the relationship between Gender diversity

and Innovation seems to be mediated by the fulfilment of the NfU at the .10 level of significance. All

other mediation models showed no significant results. Thus hypothesis 3 could therefore only be

partially confirmed.

Table 4. Results of mediation analyses

Efficacy

Performance

Innovation

R

_df

_β

_R

_df

_β

_R

_Df

_β

Step 1

0.016

(1, 30)

0.126

(1, 30)

0.126

(1, 30)

Age diversity

0.126

0.126

0.126

Step 2

0.580 ** (2, 29)

0.125

(2, 29)

0.330

(2, 29)

Age diversity

-0.347 *

0.003

0.074 *

Fuflillment of NfU

0.510 ***

0.125

0.313

ΔR

_0.564

_-0.001

_0.205

Step 1

0.171

(1, 30)

0.171 (1 , 30)

0.171

(1, 30)

Tenure diversity

0.171

0.171

0.171

Step 2

0.511 ** (2, 29)

0.147

(2, 29)

0.328

(2, 29)

Tenure diversity

-0.211

0.079

0.066

Fulfillment of NfU

0.503 ***

0.112

0.311 *

ΔR

_0.340

_-0.023

_0.158

Step 1

0.146

( 1, 30)

0.146

(1, 30)

0.146

(1, 30)

Gender diversity

0.146

0.146

0.146

Step 2

0.467 ** (2, 29)

0.125

(2, 29)

0.393 * (2, 29)

Gender diversity

-0.026

-0.005

0.228

Fulfillment of NfU

0.470 ***

0.126

0.289

ΔR

_0.322

-0.021

0.248

***. Significant at the .01 level.

**. Significant at the .05 level.

*. Significant at the .1 level.

N = 32

Hypothesis 4 stated that a work team’s fulfilment of the NfB will moderate the relationship

between the fulfilment of the NfU and the team’s performance, in such a way that by making work

team members feel unique and valued will make them perform. Which means that the relationship of

the fulfilment of the NfU and performance grows stronger under higher levels of fulfilment of the NfB,

within diverse teams.

To analyse these relationships again non-parametric resampling approaches have been used

through bootstrapping by using the process tool of Preacher and Hayes (2004). The analyses were

conducted through the PROCESS tool by Andrew Hayes (2012), model 14, which tests the conditional

indirect effect of the X on Y through Mj

= a

(b

+ b4j

V) and the direct effect of X on Y = c’

Whereas X

is the three diversity measures, Y is the three performance outcomes, Mj is the proposed mediator

NfU, and V is the proposed moderator NfB.

It appears that there is no moderated mediation of the fulfilment of the NfB on the

relationship between the fulfilment of the NfU and any of the performance indices. Although it is

notable that there are direct effects of Task interdependence (β = -.488, p < .10) and gender diversity

(β = .400, p < .10) and age diversity (β = -.803, p < .10) on Efficacy, and fulfilment of the NfB (β =.859,

p < .10) and CfI (β = -.940, p < .10) on Innovation, in the model with Age diversity as independent

variable. These direct effects imply that with higher task interdependence and with lower gender

diversity and lower age diversity feelings of efficacy grow. This effect can be explained by the amount

of cooperation needed with likewise peers and the feelings of being able to succeed in tasks. Also the

direct effects of fulfilment of NfB and CfI with innovation is noteworthy, because this implies that

when team members feel belonging to their team and/or in an inclusive climate they will become

more innovative.

Also, there are direct effects of task interdependence (β = .488, p < .05), age diversity (β =

-.803, p < .05) and gender diversity (β = -.400, p < .10) on Efficacy, and of Fulfilment of the NfU (β = .859

, p < .10) and CfI (β = -.940, p < .10) on Innovation in the model with tenure diversity as independent

variable. Also it is notable that in the model with gender diversity as independent variable, there are

significant main effects of task Interdependence (β = -.488, p < .05) gender diversity (β = .400, p < .10)

on innovation. These direct effects imply that with higher task interdependence and higher age, and

gender diversity, lower efficacy will be experienced. Even more notable is the negative relation

between CfI and innovation, suggesting that in an inclusive climate, team members will become less

innovative when there is a larger tenure diversity. For all effects, including the not-significant ones,

see figure 4 and table 5, 6 and 7.

In Hypothesis 5 it was predicted that the relationship between the fulfilment of the NfB will

be moderated by the team’s CfI in such a way that it will have a stronger moderating effect on the

relationship of the fulfilment of the NfU and its effects on performance in diverse teams.

Again, non-parametric resampling approaches have been used through bootstrapping by

using the process tool of Preacher and Hayes (2004). The analyses were conducted through the

PROCESS tool by Andrew Hayes (2012), model 18, which tests the conditional indirect effect of the X

on Y through Mj

= a

(b

+ b4j

V + b

Q + b

VQ) and the direct effect of X on Y = c’

Whereas X is the

three diversity measures, Y is the three performance outcomes, Mj is the proposed mediator NfU, V is

the proposed moderator NfB, and Q is the proposed moderator CfI.

Results show that the hypothesis appears to be only the case when the double mediated

three-way interaction effects influence the outcome variable of efficacy, contrary to the outcome

variables of team performance and innovation. As is shown in table 6, 7 and 8, and in figure 4. The

mediated three-way interaction only shows in the relationship between all diversity indices and

efficacy (age diversity: β = .264, p < .10; tenure diversity: β = .264, p < .05; and gender diversity: β =

.264, p = .05). Hypothesis 5 could therefore only be partially confirmed. It is notable that the only

significant slope differences (as can be seen in table 5) are the differences between the slope of high

fulfilment of NfB, high CfI and high fulfilment of NfB, low CfI (t = 1.986, p < .10) and high fulfilment of

NfB, low CfI, and low fulfilment of NfB, low CfI (t = -1.855, p < .10). This implicates that CfI contributes

to fulfilment of NfU and thus efficacy.

Table 5. T- and p-values for slope differences

Pair of

slopes

t-value

for slope

difference

p-value

for slope

difference

(1) and (2)

1.986 *

0.061

(1) and (3)

-0.400

0.693

(1) and (4)

0.882

0.388

(2) and (3)

-1.715

0.102

(2) and (4)

-1.855 *

0.078

(3) and (4)

1.072

0.297

*. Significant at the .10 level.

Figure 4. Simple slopes of fulfilment of the NfU predicting team efficacy for high fulfilment of the NfB

and high CfI, high fulfilment of the NfB and low CfI, low fulfilment of the NfB and high CfI, and low

fulfilment of the NfB and low CfI. With t-values and p-value for slope differences in table 5.

-3

-2

-1

0

1

2

3

Low Uniqueness

High Uniqueness

Te

am

Efficacy

(1) High Belongingness,

High CfI

(2) High Belongingness,

Low CfI

(3) Low Belongingness,

High CfI

(4) Low Belongingness,

Low CfI

Table 6. Testing all hypotheses simultaneously with predictor age diversity

Efficacy

Performance

Innovation

Variables

β

s.e.

β

s.e.

β

s.e.

Control

Team size

0.115

0.162

-0.073

0.264

0.023

0.217

Task interdependence

-0.488 *

0.222

-0.175

0.305

-0.433

0.251

Tenure diversity

0.406

0.288

0.293

0.396

0.225

0.325

Gender diversity

0.400 *

0.162

-0.072

0.264

0.355

0.217

Predictor

Age diversity

-0.803 *

0.260

-0.214

0.357

-0.197

0.293

Fulfilment of NfU

0.404

0.344

0.354

0.472

0.282

0.388

Fulfilment of NfB

0.053

0.286

0.351

0.393

0.859 *

0.323

CfI

-0.624

0.361

-0.162

0.496

-0.940 *

0.408

Fulfilment of NfU * fulfilment of NfB

-0.374

0.295

0.304

0.405

0.040

0.333

Fulfilment of NfU * CfI

0.774

0.490

0.438

0.673

-0.298

0.553

Fulfilment of NfB * CfI

-0.472

0.518

-1.049

0.712

0.141

0.584

Fulfilment of NfU * fulfilment of NfB * CfI

0.264 *

0.099

-0.167

0.136

0.344

0.111

**. Significant at the .05 level.

*. Significant at the .10 level.

N = 32

Table 7. Testing all hypotheses simultaneously with predictor Tenure Diversity

Efficacy

Performance

Innovation

Variables

β

s.e.

β

s.e.

β

s.e.

Control

Team size

0.115

0.162

-0.073

0.222

0.023

0.182

Task interdependence

-0.488 **

0.222

-0.175

0.305

-0.433

0.251

Age diversity

-0.803 **

0.260

-0.214

0.357

-0.197

0.293

Gender diversity

-0.400 *

0.192

-0.072

0.264

0.355

0.217

Predictor

Tenure diversity

0.406

0.288

0.293

0.396

0.225

0.325

Fulfilment of NfU

0.404

0.344

0.354

0.472

0.282

0.388

Fulfilment of NfB

0.053

0.286

0.351

0.393

0.859 *

0.323

CfI

-0.624

0.361

-0.162

0.496

-0.940 *

0.408

Fulfilment of NfU * fulfilment of NfB

-0.374

0.295

0.304

0.405

0.040

0.333

Fulfilment of NfU * CfI

0.774

0.490

0.438

0.673

-0.298

0.553

Fulfilment of NfB * CfI

-0.472

0.518

-1.049

0.712

0.141

0.584

Fulfilment of NfU * fulfilment of NfB * CfI

0.264 **

0.099

-0.167

0.136

0.034

0.111

**. Significant at the .05 level.

*. Significant at the .10 level.

N = 32

Table 8. Testing all hypotheses simultaneously with predictor Gender Diversity

Efficacy

Performance

Innovation

Variables

β

s.e.

β

s.e.

β

s.e.

Control

Team size

0.115

0.162

-0.073

0.222

0.023

0.182

Task interdependence

-0.488 **

0.222

-0.175

0.305

-0.433

0.251

Age diversity

-0.803 **

0.260

-0.214

0.357

-0.197

0.293

Tenure diversity

0.406

0.288

0.293

0.396

0.225

0.325

Predictor

Gender diversity

0.400 *

0.192

-0.072

0.264

0.355

0.217

Fulfilment of NfU

0.404

0.344

0.354

0.472

0.292

0.388

Fulfilment of NfB

0.053

0.286

0.351

0.393

0.859 **

0.323

CfI

-0.624

0.361

-0.162

0.496

-0.940 **

0.408

Fulfilment of NfU * fulfilment of NfB

-0.374

0.295

0.304

0.405

0.040

0.333

Fulfilment of NfU * CfI

0.774

0.490

0.438

0.673

-0.298

0.553

Fulfilment of NfB * CfI

-0.472

0.518

-1.049

0.712

0.141

0.584

Fulfilment of NfU * fulfilment of NfB * CfI

0.264 **

0.099

-0.167

0.136

0.034

0.111

**. Significant at the .05 level.

*. Significant at the .10 level.

N = 32

Discussion

Using data collected from 489 participants, in two studies and the development of a new

scale for balanced need fulfilment, the main hypotheses of this research have partially been

confirmed. It appears that under the influence of CfI, employees feel their NfB fulfilled, together

with the fulfilment of their NfU, created from more diverse work teams, they appear to experience

more efficacy. This is in line with the proposed influence of CfI by Nishii (2013) and can be seen as

the first empirical evidence for Shore’s and colleagues (2011) model of Inclusion and Diversity in

work teams.

It has been shown that, partial mediation of need fulfilment is important when it comes to

team performance. Hypothesis 1 is therefore, partially, confirmed. It appears to be important to have

the NfU and the NfB fulfilled in order to achieve higher self-rated job performance. This effect grows

to be even stronger within an inclusive climate. In line with Shore and colleagues’ (2011) predictions,

CfI enables the fulfilment of NfU and NfB in order to achieve higher performance.

It has also been shown that in teams with higher age, gender and tenure diversity, efficacy is

higher when both the NfU and NfB are fulfilled within an inclusive climate. Although efficacy can be

considered a strong predictor of performance (Stajkovic & Luthans, 1998), performance and

innovation did not seem to be higher under the influence of need fulfilment and CfI.

The presence of consistent results across the two studies with substantial methodological

differences (cross-sectional survey and an experimental design) attest to the reliability of the

findings, this suggests that the findings are not an artefact of the limitations of each type of design.

Theoretical Implications

The results of this research represent an important theoretical contribution since they

illustrate that social context can play a key role in feelings of efficacy and possible performance, as

proposed in multiple social need fulfilment theories and models (Brewer, 2003; Nishii, 2013; Shore, et

al., 2011). It is notable that all three analysed forms of diversity, be it age, tenure and/or gender, seem

to be directly influencing the fulfilment of the NfU and under the right circumstances, in combination

with fulfilment of the need for belongingness, to foster greater efficacy.

In line with past results of diversity research is has been shown that CfI is a moderating factor

that helps explain the inconsistent relationship (Cho & Mor Barak, 2008) of diversity and performance.

In addition, by using Williams and O’Reilly (1998) as a guide to the many relevant aspects on which

individuals can differ, the traditional way of defining diversity to be conceptualized as easily visible

differences such as age or gender (Hicks-Clarke & Illes, 2000; Jehn & Bezrukova, 2004), this study

shows that also less-visible diversity aspects, such as diversity of tenure, have an important role.

Although, the body of empirical testaments on how diversity and inclusive climates can

influence performance, can still be considered scarce (Nishii, 2013), this research grants more insights

on these effects.

It was expected that performance in diverse (compared to homogeneous) teams is mediated

by fulfilment of the NfU and NfB whereas lower levels of need fulfilment will lead to lower

performance in diverse work teams. Although no full mediation was found due to lack if

inter-construct correlations, partial mediation was shown, whereas performance is higher with higher need

fulfilment.

It has been hypothesized that a diverse (compared to homogeneous) work team’s CfI

moderates the strength of the relationship between team diversity and need fulfilment in such a way

that there will be more need fulfilment if there is a stronger CfI. Although no full mediation was found,

even after applying the moderator (CfI) to the model, direct effects of need fulfilment on performance

grew stronger under the influence of CfI, thus partially confirming the hypothesis.

It was predicted that within a diverse (compared to homogeneous) work team, the fulfilment

of the NfU will mediate the relationship between the diversity of a team and its performance, which

could only be partially confirmed. The lack of mediation for all constructs could lie with the fact that

correlation underlies mediation and therefore, the lack of correlation between all constructs, made

that no mediation could be found. Nonetheless, mediation of NfU was confirmed for effects of the

age, tenure and gender diversity on the efficacy, which implies that having the possibility to express

the social self and maintain a social identity leads to mere feelings of efficacy. Although efficacy can

be considered a strong predictor for performance, no significant relationship with performance could

be found. Nonetheless there was a direct relationship between higher age and tenure diversity and

innovation, supporting the assumptions of Horwitz and Horwitz (2007) on the relationship between

diversity and innovation.

It has not been shown that within a diverse (compared to homogeneous) work team, it’s

fulfilment of the NfB moderates the relationship between the fulfilment of the NfU and the team’s

performance in such a way that by making work team members feel unique and valued will make

them perform. This contradicts with the expectations sprouting from de confirmation of the second

hypothesis. Nonetheless, the direct effects within the moderated mediation model are noteworthy,

since it appears that task interdepence, age and gender diversity have a negative relationship with

efficacy when there is nog CfI. Also, contradictory to other findings, in teams with higher tenure

diversity, CfI appeared to have a negative relationship with innovation, this could imply that other

moderators (e.g. status differences (Ibarra, 1993) or power-distance derived from tenure (Shane,

1993)) have a stronger effect on innovation than just CfI.