Diversity is strategy: the effect

of R&D team diversity on

innovative performance
Marian Garcia Martinez1, Ferdaous Zouaghi2 and
Teresa Garcia Marco3
1
Kent Business School, University of Kent, UK. [Link]@[Link]
2
Public University of Navarra, Pamplona, Spain. [Link]@[Link]
3
Public University of Navarra, Pamplona, Spain. tgmarco@[Link]

Diversity in the workplace has attracted significant interest in organisations that want to
attract and retain talented employees. Breakthrough innovation requires a wider knowl-
edge base, and organisations increasingly rely on multidisciplinary R&D teams to identify
scientific developments that bridge gaps and reduce time to market. However, research on
the performance implications of R&D team diversity remains limited and the empirical
evidence inconsistent. This paper investigates the impact of surface and deep-level diver-
sity on R&D teams’ innovative performance and how diversity dimensions interact to
drive innovation. We find supportive evidence that R&D team characteristics influence
innovation outcomes, confirming our hypothesising that diversity is a valuable strategy for
an organisation to pursue as it provides greater cognitive ability. Each diversity facet how-
ever has its own distinct effects depending on the novelty of innovation and industry. Yet,
diversity is not solely positive and excessive heterogeneity could be detrimental to R&D
team performance. Our findings suggest that high diversity in gender or skills in cogni-
tively diverse teams might be negative attributes to take into consideration. Senior
managers and organisations should therefore consider the appropriate mix of capabilities
to benefit from creativity in diverse R&D teams and avoid possible conflict and distrust
associated with diversity.

1. Introduction Knippenberg et al., 2004), which can be combined in

new ways, leading to positive cognitive effects (Page,

D iversity is often described as a ‘two-edged

sword’ (Horwitz and Horwitz, 2007) or a mixed
blessing (Williams and O’Reilly, 1998) for its contra-
2007). Research even suggests that group heterogene-
ity is more important than individual ability (e.g.,
Hong and Page, 2004). On the other hand, diversity
dictory influence on organisational outcomes. Teams can reduce team performance by negatively affecting
with diverse backgrounds bring together a wider spec- cohesion, decision-making quality and members com-
trum of task-relevant knowledge, experience and per- mitment to the group (Mintzberg, 1983; Goodstein
spectives that are distinct and non-redundant (van et al., 1994).

C 2016 RADMA and John Wiley & Sons Ltd

V 311
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

This paper examines the ‘value in diversity’ The rest of the paper is structured as follows. Sec-
hypothesis (Cox et al., 1991) in innovation, an area tion 2 establishes the theoretical framework and
still largely underdeveloped in organisational innova- presents the research hypotheses. Section 3 details the
tion research (Van der Vegt and Janssen, 2003; Talke research design and methods, and Section 4 presents
et al., 2010). Despite a dramatic increase in the organ- the results of the empirical test. In Section 5, we dis-
isational use of multidisciplinary teams for innova- cuss our main findings and elaborate on their theoreti-
tion, our understanding of the influence of diversity in cal and managerial implications and present a future
innovation teams’ processes and outcomes remains research agenda on R&D team diversity, which takes
limited and the empirical evidence inconsistent into account the study’s limitations.
(Roberge and van Dick, 2010). This study aims to
address this research gap by differentiating amongst
types of diversity (i.e., gender, skills and education 2. Theoretical background
diversity – facets of surface and deep-level diversity) and hypotheses
and investigating their main effects on innovation per-
formance, hypothesising that the value in R&D team 2.1. R&D teams diversity in innovation
diversity emerges from individuals’ unique attributes
that bring different task-relevant knowledge and per- Diversity studies have examined group heterogeneity
spectives to the group (van Knippenberg et al., 2004). from two different theoretical perspectives: surface
Groups that consider a wider range of possible alter- and deep-level diversity (Jackson et al., 1995). Harri-
natives generate more creative, novel solutions (Sut- son et al. (1998, p. 97) defined surface-level diversity
ton, 2007). Further, we examine how diversity as ‘differences among group members in overt, bio-
dimensions interact to drive innovation, responding to logical characteristics that are typically reflected in
calls for a more holistic view of the overall potential physical features’. Groups are kept together by the
influence of different facets of diversity (Shore et al., perceived similarities or dissimilarities of their mem-
2009). bers’ demographic characteristics (e.g., age, gender
This paper contributes to the literature in several and race/ethnicity) (Tajfel, 1978). Surface-level diver-
distinct ways. Firstly, it makes a theoretical contribu- sity, however, generates an instant sense of dissimilar-
tion to organisational innovation research by consid- ity amongst team members, which preclude them
ering the influence of multiple dimensions of diversity from forming closer relationships (Mehra et al., 1998;
on innovation novelty – incremental and radical inno- Gibbons and Olk, 2003) and is likely to hamper social
vation. Downs and Mohr (1976) challenged the idea interaction and communication (Jehn et al., 1999).
of a single theory of innovation and argued that each Deep-level diversity refers to differences amongst
form of innovation could be explained by different group members’ psychological characteristics, such
predictive variables. Simons et al. (1999) demonstrate as cognitive abilities, attitudes, values, knowledge
the utility of treating diversity as a multifaceted con- and skills (Harrison et al., 2002). These attributes take
struct whose different facets interact with team proc- time to manifest themselves as team members need
esses to shape organisational performance in different clues from their interactions with each other to
ways. Secondly, we contribute to human capital diver- become aware of them. Compared to homogeneous
sity and innovation research by considering the inter- groups, cognitively diverse teams are superior with
action effects between diversity criteria and the regard to the prerequisites to innovation: greater
resulting impact on innovative performance. Faultline absorptive capabilities to integrate new knowledge
research shows how including individuals’ multiple and apply it to generate new ideas (Cohen and Levin-
diversity characteristics provide greater explanatory thal, 1990), higher requisite variety so they can refer
power regarding the overall influence of diversity on to their internal knowledge and expertise to solve
team performance (Lau and Murnighan, 2005). problems (Nonaka and Takeuchi, 1995) and higher
Thirdly, we contend that the industry context can con- network variety to access an external diverse knowl-
dition the effect of group diversity on performance edge pool (Katz and Tushman, 1979). Further, cogni-
(Joshi and Roh, 2009). Human resources and employ- tively diverse groups are less prone to groupthink
ees skills are key strategic assets in service innovation (Janis, 1972).
(Gallouj and Weinstein, 1997) compared to formal- A key question in innovation studies is how to
ised R&D activities in manufacturing industries maintain cohesion in cognitively diverse groups. In
(Castellacci, 2008). Hence, we argue that due to contrast to theories of social categorisation (Turner
differences in innovation processes and tasks, the et al., 1987), the interpersonal congruence approach
dynamics of R&D team diversity may differ depend- (Ely and Thomas, 2001) suggests that team members
ing on the industry setting. can achieve harmonious and effective work processes

312 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
 Diversity is strategy

Figure 1. Research framework.

by expressing rather than suppressing the attributes complement each other in R&D projects (Fenwick
that make them unique. Trying to address this contra- and Neal, 2001; Faems and Subramanian, 2013).
diction, Bernthal and Insko (1993) distinguish The management literature observes that having
between social-emotional and task-oriented cohesion women within teams improves soft management
and argue that that the latter can offset the former by skills and decision-making processes and enhances
promoting a strong analytical orientation in informa- creativity and innovation (Dessler, 2001; Bagshaw,
tion gathering and analysis, greater task focus and 2004; Egan, 2005). Women usually differ in experi-
higher attraction to work in group tasks. This paper ence and career trajectories from their male counter-
investigates the impact of surface and deep-level parts (Daily et al., 2003), but rather than equating to
diversity on innovative performance. In addition, less experience, it suggests highly diverse human and
interaction phenomena of diversity variables are con- social capital backgrounds, which might contribute to
sidered to determine the overall effects of having enhanced performance (Singh et al., 2008). Further,
social categorically diverse yet informational hetero- gender diversity may also improve the team’s external
geneous R&D teams. Our hypothesised model is relationships, allowing group members to acquire
depicted in Figure 1. knowledge and ideas through collaborations with
external groups (Joshi and Jackson, 2003), leading to
2.2. Gender diversity greater innovation capabilities. Therefore, we
hypothesise:
The diversity literature offers inconsistent results
regarding the relationship between gender diversity H1: Gender diversity is positively associated to (a)
and performance outcomes. Some scholars have incremental and (b) radical innovation
reported that gender-heterogeneous groups tend to performance.
exhibit increased conflict, low cohesion and increased
turnover (Milliken and Martins, 1996). Such findings 2.3. Skills diversity
have often been explained by social identity theory in
combination with self-categorisation theory (Turner Skills diversity refers to the portfolio of researchers,
et al., 1987). These theories suggest that greater diver- technicians and supporting staff in R&D teams
sity in salient, demographic features, such as gender, (OECD, 2002). Skills diversity usually emerges from
causes group members to employ divisive categorisa- dealing with different issues and facing diverse sets of
tions that often yield negative consequences. On the conditions, resulting in better abilities to find, inte-
other hand, several scholars have found that gender grate and use new knowledge and later developmental
diversity promotes innovation, creativity (Østergaard opportunities (Yang et al., 2009; Østergaard et al.,
et al., 2011; Dıaz-Garcıa et al., 2013; Fernandez 2011). It also affects how problems are formulated
Sastre, 2015) and productivity (Wood, 1987), with no and what types of solutions are generated (Sollner,
evidence for increased conflict (O’Reilly et al., 1998; 2010). Since innovation is an iterative process (Lund-
Pelled et al., 1999). Such positive effects of diversity vall, 1988), which requires diverse knowledge bases
have often been explained in terms of the value-in- amongst those involved in creativity and innovation,
diversity hypothesis (Cox et al., 1991), which argues knowledge diversity should generally positively affect
that differences amongst group members result in firm innovation performance (Dunbar, 1995, Dunbar,
increased information availability, perspectives, 1997). Innovation studies show that R&D teams gain
knowledge and skills (Jehn et al., 1999; Ely and significant personal experience from interactions
Thomas, 2001). With respect to gender diversity, in amongst team members with different knowledge
particular, diversity studies refer to the superior per- bases and skill sets (Chandrasekaran and Linderman,
formance of mixed teams resulting from different 2015), increasing team members’ ability to engage in
thinking styles and behavioural modes that can complex, non-routine tasks (Lin, 2014). Hence, we

C 2016 RADMA and John Wiley & Sons Ltd

V R&D Management 47, 2, 2017 313
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

expect skills diversity to be positively related to a influence of diversity on R&D teams’ performance
firm’s propensity to innovate. and have called for more complex theoretical con-
ceptualisations of diversity (Harrison and Klein,
H2: Skills diversity is positively associated to (a) 2007, van Knippenberg and Schippers, 2007). The
incremental and (b) radical innovation central premise of alignment theories is that multi-
performance. ple characteristics of individual differences are
likely to be salient at the same time, and their influ-
ence must therefore be considered simultaneously
2.4. Education diversity (Bezrukova et al., 2007). Lau and Murnighan
(1998) introduced the group ‘faultline’ concept to
Similarly, to other types of knowledge diversity, diver-
describe the configuration of teams members’
sity studies have shown inconclusive results regarding demographic attributes. The more highly correlated
the impact of education diversity on performance out- team member attributes are, the stronger faultlines
comes. By using their interpersonal dissimilarities, will be, increasing the likelihood that homogenous
educationally diverse teams are better prepared to subgroups will develop (Pelled et al., 1999; Lau and
solve complex problems because the internal pool of Murnighan, 2005). Empirical studies have shown
knowledge available to them and the integration of dif- that groups with strong faultlines are more likely to
ferent perspectives and opinions, encouraging inspir- experience a variety of negative consequences,
ing discussions, mutual learning and more novel, including greater conflict and decreased perform-
creative solutions (Jackson and Joshi, 2004; Faems ance (for a review of group faultlines refer to
and Subramanian, 2013). Similarly, Cohen and Levin- Thatcher and Patel, 2011, 2012). Increasingly, how-
thal (1990) argue that individuals’ absorptive capacity ever, faultline scholars acknowledge that faultlines
and problem-solving ability are likely to increase with may not necessarily negatively affect team func-
variety in knowledge structures based on their educa- tioning (e.g., Molleman, 2005; Bezrukova et al.,
tional background. Research has reported the influence 2009).
of education diversity in situations where teams Diverse R&D teams are built to provide high task
engage in complex cognitive tasks with multiple possi- cohesion and are characterised by heterogeneous
ble solutions (Jackson, 1996; Milliken and Martins, knowledge, skills, abilities and attitudes (Bowers
1996). However, according to social identity theory, et al., 2000). Thus, informational and social category
education diversity is likely to increase the communi- attributes co-occur in members and thus in groups
cation and coordination costs of integrating available (Jehn et al., 2008). Faultline theorising proposes dif-
knowledge or coordinating the innovation process ferential effects of particular types of faultlines; how-
(Wittenbaum and Stasser, 1996; Dahlin et al., 2005). ever, most empirical studies have focused on the
Bolli et al. (2015) argue that the benefits of stimulat- consequences of strength of faultlines in general (e.g.,
ing creativity and the costs of coordination and commu- Lau and Murnighan, 2005; Rico et al., 2012). In the
nication from high cognitive distances might differ present study, we account for the multiple faultlines
along the innovation process. Several scholars suggest that may be present within R&D teams and evaluate
that the creativity benefits are more relevant for the gen- the differential effects on firm innovation
eration of new knowledge or the invention of new prod- performance.
ucts whilst coordination and communication costs Grounded on the cognitive resource perspective,
become more important in commercialisation and mar- we hypothesise that task-related diversity (education
keting activities (e.g., Atuahene-Gima and Evangelista, and skills diversity) positively impacts group per-
2000). We hypothesise that education diversity broadens formance since members can access a wider array of
and deepens the cognitive and mental maps of R&D opinions, skills and perspectives (Horwitz and Hor-
teams, leading to superior innovative performance. witz, 2007). Group members in informational diverse
groups will engage in debates about divergent view
H3: Education diversity is positively associated to points and discuss their disagreements over group
(a) incremental and (b) radical innovation tasks, which stimulates task conflict (Jehn et al.,
performance. 1997). Task conflict is considered an important driver
of teams’ creativity and innovative behaviour (Wood-
man et al., 1993). Task-related faultlines may operate
2.5. Interaction effects between surface
as ‘healthy divides’ that stimulate effective decision-
and deep-level diversity
making processes and improve group performance
Researchers argue that considering single diversity (Gibson and Vermeulen, 2003) by utilising members’
criteria only provides limited insight on the cognitive resources. Information-based subgroups

314 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
 Diversity is strategy

may function as cohesive groups of individuals whose 3. Methodology

shared knowledge and expertise facilitates receiving
support from each other. As a results, groups may
more readily express opinions and share knowledge
3.1. Data and sample
with members of other subgroups (Nemeth and Gon- The data for the quantitative analysis have been drawn
calo, 2005); such exchanges encourage creativity and from the Technological Innovation Panel (PITEC),
healthy debate (Gibson and Vermeulen, 2003; Bezru- which is a statistical instrument for studying innova-
kova et al., 2009; Carton and Cummings, 2012), pro- tion activities of Spanish companies over time. The
moting team learning and performance (Gibson and database contains panel data for more than 12,000
Vermeulen, 2003; van Knippenberg et al., 2004). firms since 2003 and offers key advantages to the
Thus, we argue that the interaction of information study of diversity in R&D teams compared, for
diversity dimensions is positively related to innova- instance, to the Community Innovation Survey (CIS)
tion performance. where diversity is not covered. Firstly, it contains
detailed information on firms’ R&D activities and
H4: Skills diversity positively moderates the rela- more importantly classifies R&D staff in terms of
tionship between R&D team members’ education gender, education and skills, all critical variables to
diversity and innovative performance. The positive diversity research. Secondly, PITEC is designed as a
association between education diversity and inno- panel data survey so we overcome estimation prob-
vative performance increases as skills diversity lems related to the simultaneity between innovation
increases. inputs and outputs, by lagging independent variables
(Mairesse and Mohnen, 2010). Lagged-variable mod-
Faultlines are highly context dependent and might be
els have superior predictive validity, especially when
beneficial under the ‘right’ conditions (Williams and
measuring innovation outcomes (Laursen and Salter,
O’Reilly, 1998). Social-category faultlines may not
2006; Bradley et al., 2010). In addition, it mitigates
always elicit intergroup bias to the same extent; some
common method bias concerns since the time occur-
contexts may reduce the salience of social identities
rence of predictors is well defined before outcomes
and help alleviate the problems associated with fault-
are observed (Podsakoff et al., 2003).
lines (Jehn and Bezrukova, 2010). Innovation
The study was conducted using information on
involves tasks requiring frequent inter-group interac-
firms’ innovation activities and employment charac-
tions where R&D teams benefit and value differing
teristics for the years 2005–2012, including both man-
viewpoints, backgrounds and insights. We argue that
ufacturing and service firms since the literature
the innovation context favours social integration and
reports industry differences in terms of innovation
is less conducive to social categorisation, in-group
bias and intergroup conflict (Gonzalez and DeNisi, processes and tasks (e.g., Sirilli and Evangelista,
2009). Gender diversity improves the results of inter- 1998; Evangelista, 2000). For the purpose of the pres-
active decisions (Fenwick and Neal, 2001), broadens ent research, the dataset was confined to companies
the perspective and contributes to better social rela- with a positive expenditure in internal R&D for at
tions and an open work climate and debate (Nielsen least 1 year during the studied period1. Table 1 shows
and Huse, 2010a,b). Hence, innovation provides a the main features of the sample.
context in which demographic faultlines can be a
‘healthy divide’ (Gibson and Vermeulen, 2003; Iseke 3.2. Measures
et al., 2015).
3.2.1. Dependent variable
H5: Gender diversity positively moderates the Innovative performance is the dependent variable of
relationship between R&D team members’ educa- the model measured as the percentage of the firm’s
tion diversity and innovative performance. The total sales from innovations (Hitt et al., 1996). Con-
positive association between education diversity sistent with CIS-based studies (e.g., Laursen and
and innovative performance increases as gender Salter, 2006; Sofka and Grimpe, 2010), we distinguish
diversity increases. between incremental and radical innovation depend-
ing on their newness to the company or the market
H6: Gender diversity positively moderates the rela- place. Radical innovation is measured as the percent-
tionship between R&D team members’ skills diver- age of the firm’s total sales from innovations new to
sity and innovative performance. The positive the market in the last 2 years. Incremental innovation
association between skills diversity and innovative is defined as the percentage of the firm’s total sales
performance increases as gender diversity increases. from innovations new to the firm in the last 2 years.

C 2016 RADMA and John Wiley & Sons Ltd

V R&D Management 47, 2, 2017 315
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

Table 1. Samples descriptive statistics

Variables Characteristics Manufacturing firms (%) Service firms (%)

Gender Male 74.2 69.8

Female 25.8 30.2
Education PhD 5.1 8.1
Bachelor 42.7 56.7
Secondary education 22.8 19.8
Other studies 29.4 15.4
Skills Researchers 43.1 51.5
Technicians 42.5 41.6
Supporting staff 14.4 6.9
Number of employees (size) Less than 50 44.2 61.9
Between 50 and 99 20 11.3
Between 100 and 449 28.7 16.9
500 or more 7.1 9.9
R&D team size Less than five members 47.7 43.8
Between 5 and 19 members 39.6 36.9
Between 20 and 39 members 7.4 9.5
40 or more 5.3 9.8

This lag ensures that sufficient time has elapsed since based on three different job functions related to the
the introduction of innovations and their performance expertise of R&D team members: (1) researchers, (2)
evaluation (Langerak et al., 2008). Past studies have technicians and (3) supporting staff (OECD, 2002).
shown that this self-report subjective performance Blau’s index for skills diversity can vary from 0
measure is reliable and correlates highly with other (when there is only one skills area represented in the
objective indicators of innovation performance R&D team) to 0.68 (when there is an equal number of
(Hagedoorn and Cloodt, 2003). R&D members across all three skills areas). We struc-
tured education into four categories: (1) PhD, (2)
Bachelor, (3) Secondary education and (4) other stud-
3.2.2. Independent variables ies (i.e., vocational training, etc.). In this case, Blau’s
R&D team diversity is defined as the distribution of index varies from 0 (when all R&D team members
differences amongst R&D team members of the firm fall within the same educational level) to 0.75 (when
with respect to a common attribute (Harrison and there are equal numbers of R&D team members
Klein, 2007). In line with team diversity studies using across all educational levels).
categorical diversity attributes, we use Blau’s (1977)
index of heterogeneity:
3.2.3. Control variables
X
k
Firm size has been related to innovation capabilities
D512 p2i
i51
and the novelty of innovations (Ettlie and Rubenstein,
1993; Chandy and Tellis, 1998, 2000). To account for
where k represents the total number of categories of a the non-normality of the size measure, a logarithm
variable, and pi is the proportion of R&D team mem- transformation was used (Damanpour, 1992). In addi-
bers who fall in category k. The minimum value of 0 tion, we account for non-linear effects of firm size by
occurs when all R&D team members fall within the computing firm size squared (Acs and Audretsch,
same category and there is no variety (e.g., all R&D 1990, 1991).
team members are female). The greater the distribu- R&D team size: the literature in work groups notes
tion of characteristics across categories, the higher the that team size is a key variable influencing group
diversity index is. dynamics and performance (Pelled et al., 1999; Sethi
We included three diversity measures: gender, et al., 2001). Large teams have more potential for het-
skills and education. Since there are only two gender erogeneity and this can affect the type of innovation
categories (male and female), Blau’s gender diversity developed by the company (Lopez Cabrales et al.,
index can take on values ranging from 0 (when a 2008). R&D team size is measured by (the natural
R&D team is dominated by a single gender) to 0.5 logarithm of) the number of full-time employees in
(when there is a balanced number of men and women the R&D department. We account for non-linear
in the R&D team). The categorisation of skills was effects of R&D team size by computing the squared

316 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
 Diversity is strategy

term. As team size grows, the complexity of the com- their respective operational maximum (Spickermann
munication structure between members increases dra- et al., 2014). Table 2 indicates that our sample shows
matically (Zenger and Lawrence, 1989) leading to comparably adequate diversity across the three
greater coordination costs (Hoegl, 2005). dimensions included in the study.
Innovation Intensity is defined as the innovation Table 3 provides the descriptive statistics of each of
expenditure share of sales (de Faria et al., 2010). Inno- the variables. Correlation values amongst all variables
vation expenditure includes internal and external are generally low to moderate, suggesting low colli-
R&D, acquisition of machinery and acquisition of nearity risks. The highest correlation is 0.55, far less
knowledge for innovation, training for innovation and than the problematic level (0.75) (Tsui et al., 1995).
preparation of the market for the introduction of inno- This is confirmed by the analysis of Variance of Infla-
vations. Further, we account for a non-linear relation- tion (Vif). The maximum Vif value is 1.58, well
ship between innovation intensity and innovative below the rule of thumb cut-off of 10, which again
performance by computing the squared term. The indicates that there were no serious multicollinearity
impact of innovation intensity on innovative perform- problems in the models (Neter et al., 1996).
ance decreases when firms proportionally allocate As predicted, the correlation matrix shows that
more resources to innovation due to the difficulties to incremental innovation is significantly correlated with
integrate and apply excessive knowledge as greater gender (r 5 0.05, P < 0.01), skills (r 5 0.05,
information and control systems are required (Dyer P < 0.01) and education diversity (r 5 0.06,
and Singh, 1998). P < 0.01). However, radical innovation is signifi-
Technology Intensity: firms’ innovation behaviour cantly correlated with gender (r 5 0.02, P <0.01) and
is related to their industry affiliation (Audretsch, education diversity (r 5 0.01, P < 0.01), but not with
1997; Malerba et al., 1997); hence, we controlled for skills diversity (r 5 0.01, n.s.). Faultline research
industry effects following the OECD classification of shows that the alignment of diversity characteristics
would result in the formation of subgroups (Lau and
industries in terms of technology intensity and knowl-
Murnighan, 2005). Hence, the correlations of the vari-
edge intensity (OECD, 2005). We created four indus-
ous diversity dimensions included in this study might
try dummies to identify manufacturing firms
account for certain interactions between the individual
belonging to high-tech, medium-high, medium-low
measures concerning innovation output estimates.
and low-tech industry and two dummy variables for
We use Tobit regression models to test the main
service industries: knowledge-intensive business and
effects of diversity facets on innovative performance
low knowledge-intensive business services. We used
(H1–H3). The dependent variables (radical and incre-
the high-tech industry as the baseline for manufactur-
mental innovation performance) are percentage meas-
ing models and the knowledge-intensive services for ures and thereby conditioned on values between 0%
service models. and 100%. As the distribution of radical and incre-
Year effects: we used firm-level innovation per- mental innovation is highly skewed to the left, the
formance data from 2005 to 2012; hence, 8-year dum- assumption of a normal distribution of the residuals
mies variables were included to control unobserved made in a Tobit analysis is violated (significance of
factors that change over time but remain relatively Shapiro–Wilk test of 0.000 for both dependent varia-
constant across industries. bles). Thus, we use the logarithmic transformation of
the Tobit model (Filippucci et al., 1996; Papalia and
3.3. Method and data analysis Di Iorio, 2001). In addition, we lagged all explanatory
and control variables (except for industry dummies
To test that all three diversity criteria show an which do not vary across panel waves) by one period,
adequate level of heterogeneity and are comparable in consistent with the survey implementation rhythm, to
their level of diversity, we normalised our diversity avoid simultaneity and reverse causality problems
indices on a 0–1 metric scale by dividing them by (Mairesse and Mohnen, 2010). This reduced our sam-
ple to an unbalanced panel of seven years and 35,107
Table 2. Standardised diversity indices (Blau’s index of observations.
heterogeneity)
To test the interaction effects (H4–H6), we used
Full Manufacturing Service hierarchical regression analysis (Frazier et al., 2004).
sample firms firms To prevent multi-collinearity problems between the
Gender 0.46 0.42 0.52 main effects and the interaction effects, all diversity
Skills 0.51 0.52 0.43
measures were mean-centred before calculating the
interaction terms (Aiken and West, 1991), and subse-
Education 0.49 0.55 0.44
quently checked to ensure that all Vif values were

C 2016 RADMA and John Wiley & Sons Ltd

V R&D Management 47, 2, 2017 317
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

Table 3. Means, standard deviations and correlation coefficients

Variable Mean S.D. 1 2 3 4 5 6 7 8

1. Radical innovation 13.99 25.78 1

2. Incremental innovation 40.44 42.96 20.16** 1
3. Gender 0.23 0.20 0.02** 0.05** 1
4. Education 0.37 0.24 0.02** 0.06** 0.25** 1
5. Skills 0.34 0.23 0.01 0.05** 0.21** 0.55** 1
6. Firm size (Ln) 4.10 1.56 20.07** 0.08** 0.21** 0.22** 0.15** 1
7. R&D team Size (Ln) 1.29 1.37 0.11** 0.08** 0.37** 0.42** 0.35** 0.49** 1
8. Inn intensity 0.18 0.52 0.12** 20.04** 0.09** 0.02** 20.01 20.24** 0.16** 1
Vif 1.42 1.43 1.40 1.29 1.30 1.27 1.22 1.35
S.D., standard deviation; Vif, Variance Inflation Factor.
*P < 0.05, **P < 0.01.

below 10 (Neter et al., 1996). The Tobit regression decrease the probability of introducing incremental
models show an overall adequate level of validity and radical innovations.
according to various statistics commonly used for Models 2 and 8 (manufacturing – Table 4) and
interpretation (Hair et al., 2010): highly significant Models 14 and 20 (service – Table 5) tested H1 (gen-
model chi-squares; further, the smaller values of the der), H2 (skills) and H3 (education), that diversity
Akaike information criterion and the Bayesian infor- variables are positively associated to incremental and
mation criterion in models compared with each previ- radical innovation performance. A diverse gender
ous model suggest that the relative goodness of fit in composition of R&D teams is positively associated
each model improved significantly on the previous with radical (manufacturing and service) and incre-
one. The Tobit regression models were estimated mental (manufacturing) innovation. Hence, H1 is
using STATA 13. largely accepted. Skills diversity influences radical
(manufacturing) and incremental (manufacturing and
service) innovation. Thus, H2 is largely accepted.
4. Results Diversity in education exhibits the same impacts as
gender diversity; however, it holds the strongest
4.1. Main effects of diversity on innovative effect of all diversity measures for radical innovation.
performance Consequently, H3 is largely accepted. These findings
corroborate our premise that diversity is a multidi-
Models 1 and 7 (manufacturing – Table 4) and Mod- mensional construct that impacts innovative
els 13 and 19 (service – Table 5) present the effects of performance differently depending on the novelty of
the control variables. R&D team size has a significant innovation and industry context. In manufacturing
and positive effect on the likelihood of introducing sectors, both surface and deep-level diversity posi-
both radical and incremental innovation; however, the tively impacts performance outcomes although educa-
quadratic effect is negative indicating that additional tion and skills attributes that form a team’s cognitive
increases in R&D team size will reduce the probabil- resource base have the strongest effect on radical
ity of additional radical and incremental innovations. innovation. In contrast, diversity has a more limited
Innovation intensity has a significant and positive impact in the service sector with distinct impacts
effect on radical innovation, as expected, since radical depending on the novelty of innovation.
innovation embodies new knowledge so greater inno-
vation support is required; however, the quadratic
effect is negative indicating that additional increases 4.2. Moderation effects
in innovation expenditure will reduce the probability H4 predicts that skills diversity positively moderates
of radical innovation. Firm size also has a significant the education diversity-innovative performance link.
and positive effect on innovative performance for all Contrary to our hypothesised relationship, the interac-
models but Model 19; the quadratic effect is negative tion term for skills diversity 3 education diversity is
for incremental innovation suggesting that although significant but negative for all models but Model 10
larger firms tend to introduce more incremental inno- (radical manufacturing innovation). A simple slope
vations than smaller firms, the oversize can generate analysis (Aiken and West, 1991; Dawson, 2014)
monitoring costs and management problems that shows that, at low levels of R&D team skills diversity,

318 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
V
Table 4. Results of Tobit regression analysis: effect of diversity on firm innovation performance (manufacturing firms)
Incremental innovation Radical innovation

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10 Model 11 Model 12
B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.)

R&D team diversity

H1: gender 0.335*** 0.335*** 0.323*** 0.337*** 0.323*** 0.132** 0.135** 0.124* 0.143** 0.137**
(0.063) (0.063) (0.063) (0.063) (0.063) (0.064) (0.064) (0.064) (0.064) (0.064)
H2: skills 0.291** 0.291** 0.251*** 0.287*** 0.255** 0.213** 0.194** 0.186** 0.194** 0.178**
(0.087) (0.087) (0.088) (0.087) (0.089) (0.088) (0.088) (0.089) (0.089) (0.089)
H3: education 0.183* 0.184* 0.134* 0.178* 0.136 0.561*** 0.544*** 0.525*** 0.531*** 0.513***
(0.094) (0.095) (0.097) (0.096) (0.097) (0.096) (0.096) (0.098) (0.097) (0.098)

C 2016 RADMA and John Wiley & Sons Ltd

Moderating effects
H4: skills 3 20.543** 20.583** 20.389 20.211
education (0.235) (0.249) (0.238) (0.253)
H5: gender 3 0.012 0.154 20.352* 20.105
education (0.192) (0.228) (0.195) (0.232)
H6: gender 3 skills 20.089 20.028 20.464** 20.361
(0.182) (0.220) (0.185) (0.224)
Control variables
Firm size 0.322*** 0.252*** 0.252*** 0.248** 0.251*** 0.249** 0.062* 20.019 20.023 20.022 20.023 20.025
(0.031) (0.032) (0.032) (0.032) (0.032) (0.032) (0.031) (0.033) (0.033) (0.033) (0.033) (0.033)
Firm size Sq 20.164*** 20.141*** 20.141*** 20.140*** 20.141*** 20.141*** 0.014 0.041** 0.043** 0.042** 0.043** 0.043**
(0.019) (0.019) (0.019) (0.019) (0.019) (0.019) (0.018) (0.019) (0.019) (0.019) (0.019) (0.019)
R&D team Size 0.178*** 0.129** 0.129** 0.135** 0.163*** 0.168** 0.346*** 0.296*** 0.299*** 0.300*** 0.378*** 0.381***
(0.041) (0.041) (0.041) (0.041 (0.051) (0.052) (0.040) (0.040) (0.040) (0.040) (0.051) (0.051)
R&D team 20.007** 20.005** 20.005** 20.006** 20.009** 20.009** 20.012*** 20.010*** 20.010*** 20.010*** 20.017*** 20.017***
Size Sq (0.002) (0.002) (0.002) (0.002) (0.004) (0.003) (0.002) (0.002) (0.002) (0.002) (0.003) (0.003)
Innovation 0.213 0.004 0.004 20.001 0.002 0.002 0.428*** 0.384*** 1.821*** 1.827*** 1.819*** 1.817***
intensity (0.205) (0.205) (0.205) (0.205) (0.205) (0.205) (0.040) (0.001) (0.196) (0.195) (0.196) (0.196)
Innovation 20.073* 20.401 20.040 20.039 20.039 20.039 20.014*** 20.012*** 20.267*** 20.268*** 20.273*** 20.267***
intensity Sq (0.042) (0.042) (0.042) (0.042) (0.042) (0.042) (0.001) (0.001) (0.038) (0.038) (0.038) (0.038)
Intercept 0.845*** 0.448*** 0.845*** 0.880*** 0.872*** 0.904*** 20.100 20.580*** 20.190** 20.176** 20.132 20.118
(0.086) (0.100) (0.088) (0.089) (0.087) (0.089) (0.087) (0.102) (0.089) (0.090) (0.088) (0.089)

R&D Management 47, 2, 2017

Log-likelihood 238,222.10 238,184.69 238,184.68 238,182.026 238,184.57 238,181.77 233,973.29 233,928.08 233,926.46 233,926.75 233,924.95 233,924.45
v2 4,532.92*** 4,607.74*** 4,607.74*** 4,613.07*** 4,607.98*** 4,613.58*** 488.72*** 579.15*** 582.39*** 581.81*** 585.40*** 586.41***
Diversity is strategy

319
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

there is a significantly positive relationship between

68,079.32
Model 12
educational diversity and incremental innovation

67,894.9
B(S.E.)
(manufacturing: b 5 0.315, P < 0.05; service:
b 5 0.380, P < 0.05), whilst at high levels of skills
diversity, the relationship between educational diver-

67,891.91
68,060.29
Model 11
B(S.E.) sity and incremental innovation is not statistically sig-
nificant (manufacturing: b 5 20.046, n.s.; service:
b 5 20.144, n.s.; Figures 2a and 2b). The differences
between the slopes were significant (manufacturing: t
67,895.51
68,063.88
Model 10

value 5 22.90, P < 0.05; service: t value 5 22.81;

B(S.E.)

P < 0.05). Regarding radical innovation in the service

sector, Figure 2c shows that the positive effect of
R&D team education diversity on radical innovation
67,894.92
68,063.3
Model 9

increases more slowly as skills diversity increases.

B(S.E.)

This result indicates that skills diversity weakens the

positive effect of education diversity. The difference
between the slopes was significant (t value 5 22.31,
67,896.16
68,056.52
Model 8

P < 0.05).
B(S.E.)
Radical innovation

The interaction term for gender 3 education diver-

sity (H5) is also significantly negative for radical man-
ufacturing innovation (Model 9 in Table 4,
67,980.6
68,116.9
Model 7

b 5 20.352, P < 0.1), further the main effect is posi-

B(S.E.)

tive (Model 7 in Table 4, b 5 0.132, P < 0.05). In

service firms, gender diversity negatively moderates
the effect of education diversity on incremental inno-
76,409.54
76,593.95
Model 6

vation (Model 15 in Table 5, b 5 20. 808, P < 0.05),

B(S.E.)

which remains consistent in the full model (Model 18,

b 5 20. 681, P < 0.05). It should be noted that the
main effect of gender diversity is not significant
76,411.14
76,579.52
Model 5

(Model 13 in Table 5, b 5 0.149, P > 0.10). Figure 3a

S.E., standard error; AIC, Akaike information criterion; BIC, Bayesian information criterion.
B(S.E.)

Year and sector dummy variables were included in the analysis but results are omitted here.

shows that the positive effect of educational diversity

on radical manufacturing innovation increases more
slowly when gender diversity increases. The differ-
76,406.05
76,574.43

ence between the slopes was significant (t val-

Model 4
B(S.E.)

ue 5 22.27, P < 0.05). For service firms, at high

levels of gender diversity, the relationship between
educational diversity and incremental innovation is
76,411.38
76,579.75

not significant (b 5 20.173, n.s.) whereas at low lev-

Model 3
B(S.E.)

els of gender diversity, the relationship is positive and

significant (b 5 0.479, P < 0.005) (Figure 3b). The
difference between the slopes was significant (t val-
ue 5 23.63, P < 0.001). Consequently, H5 is not
76,409.38
76,569.74
Incremental innovation

Model 2
B(S.E.)

supported.
We found similar results for the interaction term
for gender 3 skills diversity (H6). Figure 4a shows a
*P < 0.1, **P < 0.05, ***P < 0.001.

negative moderating effect for radical manufacturing

76,614.51
76,478.2
Model 1
B(S.E.)

innovation with a positive relationship between R&D

team skills diversity and radical innovation for low
levels of gender diversity (b 5 0.381, P < 0.001)
Table 4. (Continued)

whereas the relationship is not statistically significant

for high levels of gender diversity (b 5 0.0007, n.s.).
The difference between the slopes was significant
(t value 5 22.28, P < 0.05). For service firms (Figure
4b), at high levels of gender diversity, the relationship
AIC
BIC

between skills diversity and incremental innovation is

320 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
V
Table 5. Results of Tobit regression analysis: effect of diversity on firm innovative performance (service firms)
Incremental innovation Radical innovation

Model 13 Model 14 Model 15 Model 16 Model 17 Model 18 Model 19 Model 20 Model 21 Model 22 Model 23 Model 24
B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.) B(S.E.)

R&D team diversity

H1: gender 0.149 0.118 0.129 0.145 0.108 0.217** 0.219** 0.202** 0.216** 0.210**
(0.100) (0.101) (0.101) (0.100) (0.101) (0.092) (0.092) (0.092) (0.092) (0.093)
H2: skills 0.257** 0.224* 0.268** 0.257** 0.238* 0.061 0.064 0.072 0.062 0.084
(0.130) (0.131) (0.130) (0.131) (0.131) (0.119) (0.119) (0.119) (0.119) (0.119)
H3: education 0.159 0.152 0.117 0.132 0.121 0.521*** 0.522*** 0.490*** 0.513*** 0.482***
(0.137) (0.137) (0.139) (0.138) (0.139) (0.126) (0.126) (0.127) (0.126) (0.128)

C 2016 RADMA and John Wiley & Sons Ltd

Moderating effects
H4: skills 3 20.776** 20.572 20.584* 20.588*
education (0.370) (0.385) (0.338) (0.352)
H5: gender 3 20.808** 20.681** 0.073 0.279
education (0.285) (0.342) (0.261) (0.314)
H6: gender 3 skills 20.536* 20.064 20.164 20.195
(0.274) (0.333) (0.50) (0.305)
Control variables
Firm size 0.178*** 0.148*** 0.143*** 0.144*** 0.126*** 0.125*** 20.050 20.098** 20.097** 20.101** 20.105** 20.107***
(0.037) (0.038) (0.038) (0.038) (0.036) (0.037) (0.033) (0.034) (0.035) (0.035) (0.033) (0.034)
Firm size Sq 20.062*** 20.051** 20.050*** 20.051** 20.051*** 20.049** 20.032** 20.016 20.016 20.015 20.016 20.015
(0.015) (0.015) (0.015) (0.015) (0.015) (0.015) (0.013) (0.014) (0.014) (0.014) (0.014) (0.014)
R&D team size 0.142*** 0.125** 0.134*** 0.128** 0.130*** 0.136*** 0.323*** 0.296*** 0.295*** 0.298*** 0.298*** 0.297***
(0.037) (0.037) (0.037) (0.037) (0.038) (0.037) (0.033) (0.338) (0.033) (0.033) (0.034) (0.033)
R&D team 20.004** 20.004 20.004** 20.004** 20.004*** 20.004** 20.008*** 20.007*** 20.007*** 20.007*** 20.007*** 20.007***
size Sq (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Innovation 20.036 20.119 20.104 20.109 20.112 20.098 0.854*** 0.734*** 0.733*** 0.741*** 0.737*** 0.739***
intensity (0.120) (0.122) (0.122) (0.122) (0.122) (0.122) (0.108) (0.109) (0.110) (0.110) (0.110) (0.110)
Innovation 20.019 20.007 20.009 20.008 20.008 20.010 20.128*** 20.110*** 20.111*** 20.112*** 20.111*** 20.112***
intensity Sq (0.023) (0.023) (0.023) (0.023) (0.023) (0.023) (0.021) (0.021) (0.021) (0.021) (0.021) (0.021)
Intercept 20.121** 20.367** 20.096 20.068 20.083 20.046 20.535*** 20.889*** 20.521*** 20.489*** 20.535*** 20.508***
(0.099) (0.120) (0.099) (0.100) (0.099) (0.101) (0.091) (0.111) (0.091) (0.093) (0.062) (0.094)

R&D Management 47, 2, 2017

Log-likelihood 220,865.23 220,858.14 220,854.12 220,855.94 220,856.22 220,852.91 219,662.52 219,644.13 219,644.09 219,642.64 219,643.92 219,642.23
v2 1,887.30*** 1,901.48*** 1,909.51*** 1,905.87*** 1,905.32*** 1,911.94*** 480.46*** 517.25*** 517.33*** 520.22*** 517.68*** 521.06***
Diversity is strategy

321
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

not significant (b 5 0.040, n.s.) whereas at low levels

Model 24

39,326.46
39,482.71
B(S.E.)
of gender diversity, the relationship is positive and
significant (b 5 0.474, P < 0.001). The difference
between the slopes was significant (t value 5 21.79,
Model 23 P < 0.1). Hence, H6 is not supported. Taken together,

39,325.84
39,467.21
B(S.E.)
these findings indicate that increasing levels of gender
diversity lead to the formation of homogeneous sub-
groups in R&D teams, creating as a result social bar-
riers and constituting a principal impediment to group
Model 22

39,323.29
39,464.66
B(S.E.)

cohesion (Blau, 1977). When group cohesion is

undermined, group performance suffers.
Model 21

4.3. Robustness checks

39,326.19
39,467.56
B(S.E.)

Several robustness checks have been carried out to

assess the sensitivity of the results to changes in
model specifications. We estimated our model using
Model 20

39,324.27
B(S.E.)

39,458.2

Ordinary Least Squares (OLS) and Poisson regression

Radical innovation

and the results were consistent. Additionally, we

applied an Ordered Probit model similar to Henkel
(2006) where the dependent variable can take values
Model 19

39,355.05
39,466.66
B(S.E.)

between 1 and 5 (‘1’ indicates that the share lies in the

first quartile (0%–20%), ‘2’ between 21%–40%, etc.).
This model specification allows for a non-linear
dependence of the share of sales from radical and
41,747.82
41,904.07
Model 18
B(S.E.)

incremental innovation on the explanatory variables

inside the interval (0%–100%]. The results were
highly robust to these changes in specification.
Model 17

41,750.45
41,891.82
B(S.E.)

S.E., standard error; AIC, Akaike information criterion; BIC, Bayesian information criterion.
Year and sector dummy variables were included in the analysis but results are omitted here.

5. Discussion and conclusions

Our aim in this research has been to add to our under-
Model 16
B(S.E.)

41,891.27

standing of the performance effects of R&D team

41,749.9

composition. The multiple and often contradictory

group dynamics that emerge in heterogeneous teams
present organisations with major challenges (Post
Model 15

et al., 2009). Specifically, we hoped to demonstrate

B(S.E.)

41,746.25
41,887.62

the significance of the specific context/task in which

different facets of diversity impact R&D team proc-
esses and outcomes (Haas, 2010). Further, we provide
Model 14

support to the call for a more holistic view of the over-

B(S.E.)

41,752.28
41,886.21
Incremental innovation

all influence of team heterogeneity on innovativeness

(Shore et al., 2009).
We find supportive evidence that R&D team com-
*P < 0.1, **P < 0.05, ***P < 0.001.

position influence innovative performance; however,

41,760.47
41,872.08
Model 13
B(S.E.)

the effect of diversity differs in terms of the dimension

considered and the novelty of innovation (Lopez
Cabrales et al., 2008). Further, our findings show dif-
Table 5. (Continued)

ferent impacts for manufacturing and service indus-

tries, supporting the view that these sectors differ
fundamentally in innovation processes and tasks,
which, in turn, affect their human resource practices
(Jackson and Schuler, 1995; Joshi and Roh, 2009). In
AIC
BIC

manufacturing firms, both surface and deep-level

322 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
 Diversity is strategy

Figure 2. Moderating role of skills diversity on the education diversity-innovation performance relationship.

diversity positively impacts performance outcomes education diversity, whereas gender diversity is more
whereas diversity has a more limited impact in the relevant for incremental innovation in manufacturing
service sector with distinct impacts depending on the firms. These results correspond with Jehn et al.
novelty of innovation. (1999), who reported a positive relationship between
Our results show positive effects of information information diversity and team performance in high
diversity on innovative performance. Sharing, proc- complex tasks.
essing and integration of diverse opinions and Despite recent theorising that faultlines, and spe-
approaches to problem solving on the basis of R&D cially task-related faultlines might enhance creativity
team members’ own area of expertise and background (e.g., Nishii and Goncalo, 2008; Bezrukova and
positively impact innovation novelty. Findings sug- Uparna, 2009), our moderation results suggest that
gest that though diversity (i.e., variety of education excessive diversity could lead to less than desirable
and skills) (Post et al., 2009) in R&D teams has the outcomes in R&D teams. R&D team members’ edu-
strongest effect on radical innovation, in particular cation is tightly linked to their gender and skills,

Figure 3. Moderating role of gender diversity on the education diversity-innovation performance relationship.

C 2016 RADMA and John Wiley & Sons Ltd

V R&D Management 47, 2, 2017 323
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

Figure 4. Moderating role of gender diversity on the skills diversity-innovation performance relationship.

leading to the formation of subgroups created by heterogeneous profile; however, the influence of each
aligned characteristics (Gibson and Vermeulen, diversity facet varies depending on the novelty of
2003). As subcategories develop within R&D teams, innovation and industry. Hence, careful consideration
group cohesion is undermined and group performance should be given to the appropriate mix of capabilities
suffers (Faems and Subramanian, 2013). R&D team rather than the absolute level of capabilities.
members’ open expression of their distinctive per- Interaction effects in particular suggest that too
spectives and expertise might generate conflict and much heterogeneity could be detrimental to firm inno-
distrust amongst R&D team, affecting team vation performance. Diversity-oriented HRM prac-
performance. tices can capitalise on the benefits from diverse R&D
teams by reducing the potential for misunderstanding
and conflict. Training and facilitation in conflict reso-
5.1. Contributions and managerial lution and negotiation would enable R&D team mem-
implications bers to integrate diversity of perspectives and
Extending prior diversity research (e.g., Jackson and information into the group (Ancona and Caldwell,
Joshi, 2004) and consistent with a growing body of 1992). Training team members to understand the per-
evidence from research on group faultline (e.g., Gib- spective of others helps group formation (Williams
son and Vermeulen, 2003; Bezrukova et al., 2009), et al., 2007). Beyond individual interventions, HRM
this papers enhances our understanding of the per- practices should also focus on systems and processes,
formance implications of R&D team diversity by con- including organisational rewards for team collective
sidering multiple diversity facets simultaneously (Lau performance, which might help mitigate surface-level
and Murnighan, 1998). The fragmentation effects differences and encourage greater cooperation (Harri-
from increasing diversity underscore the importance son et al., 2002; Swanson and Holton, 2009).
of going beyond merely increasing R&D team diver-
sity to focusing how surface-level and deep-level
5.2. Limitations and future research
diversity facets interact and how diversity can be
managed to avoid possible conflict and distrust. We acknowledge several limitations in our paper and
Our study provides managers with valuable insights suggest related opportunities for future research.
on how diversity can be used as a valuable human Firstly, this paper examines the influence of diversity
capital to foster creativity and innovation. R&D in R&D teams on innovative performance. Future
team configuration is within the control of the firm, as diversity studies should expand beyond R&D teams to
such, firms have the flexibility and discretion to consider heterogeneity related to various dimensions
encourage or discourage diversity (Auh and Menguc, and levels within the firms, including interactions with
2005). Overall, all three diversity facets positively top management in innovation processes (Harrison
impact firm innovation performance, confirming our et al., 1998). Secondly, our findings suggest that R&D
hypothesising that diversity is a valuable strategy as it teams’ education diversity is tightly coupled with gen-
provides greater cognitive diversity. Innovativeness der and skills diversity, with a resulting detrimental
is enhanced when R&D team members exhibit a impact on team performance. Future research should

324 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
 Diversity is strategy

therefore seek to expand our research by investigating exploratory study of marketing and R&D perspectives.
the moderating role of R&D organisational capabil- Management Science, 46, 1269–1284.
ities (i.e., processes, roles and systems) to leverage Audretsch, D.B. (1997) Technological regimes, industrial
diversity as a strategic resources to foster innovation demography and the evolution of industrial structures.
and creativity. Thirdly, future studies should explore Industrial and Corporate Change, 6, 49–82.
Auh, S. and Menguc, B. (2005) Top management team
the potential curvilinear relationship between R&D
diversity and innovativeness: the moderating role of
team diversity and performance outcomes to better
interfunctional coordination. Industrial Marketing Man-
understand the complexities surrounding team diver- agement, 34, 249–261.
sity (Horwitz and Horwitz, 2007). Fourthly, our data Bagshaw, M. (2004) Is diversity divisive? A positive train-
do not capture information on how R&D teams are ing approach. Industrial and Commercial Training, 36,
structured. Different R&D structures (i.e., centralised 153–157.
vs autonomous R&D teams linked to SBUs) might Bernthal, P.R. and Insko, C.A. (1993) Cohesiveness
activate or inhibit certain group faultlines resulting in without groupthink: the interactive effects of social
different R&D team performance. Finally, we use and task cohesion. Group & Organization Manage-
data from Spain, where R&D team diversity is rather ment, 18, 66–87.
limited, particularly regarding female participation in Bezrukova, K., Jehn, K.A., Zannutto, E.L., and Thatcher,
science and technology (Bordons et al., 2009; S.M.B. (2009) Do workgroup faultlines help or hurt? A
Maule on and Bordons, 2014). Hence, our data do not moderated model of faultlines, team identification, and
group performance. Organization Science, 20, 35–50.
exhibit an equal distribution of firms across all diver-
Bezrukova, K., Thatcher, S.M.B., and Jehn, K. (2007)
sity variables. Evidence from other countries on the
Group heterogeneity and faultlines: comparing alignment
differential effect of group heterogeneity on innova- and dispersion theories of group composition. In: Behfar,
tive performance might help to develop more general K.J. and Thompson, L.L. (eds), Conflict in Organiza-
empirical evidence. Specifically, future studies should tional Groups: New Directions in Theory and Practice.
consider the influence of country-specific dimensions Evanston, IL: The Northwestern University Press, pp.
(i.e., culture and institutions), in innovation perform- 57–92.
ance in cross-country studies. Bezrukova, K. and Uparna, J. (2009) Group splits and cul-
ture shifts: a new map of the creativity terrain. In: Neale,
M.A., Mannix, B. and Goncalo, J. (eds), Research on
Acknowledgements Managing Groups and Teams. Stamford, CT: JAI Press,
pp. 161–191.
The authors gratefully acknowledge the funding Blau, P.M. (1977) Inequality and Heterogeneity: A Primi-
tive Theory of Social Structure. New York, NY: The
received through the AGL2015-65897-C3-1_R
Free Press.
research project, and they would also like to express
Bolli, T., Renold, U., and W€orter, M. (2015). Vertical Edu-
their gratitude for the financial support received from cational Diversity and Innovation Performance. KOF
the Public University of Navarra for the acquisition of Working Papers No. 395. Zurich.
the pre-doctoral scholarship (Modality type B). A Bordons, M., Mauleon, E., Gomez Caridad, I., Morillo, F.,
word of sincere appreciation to the two anonymous Sancho, R., Fernandez, M.T., Garcia, C., and Arias, E.
referees for their valuable comments and suggestions. (2009). Indicadores de actividad tecnologica desagrega-
dos por sexo. Madrid: Instituto de la Mujer.
Bowers, C.A., Pharmer, J.A., and Salas, E. (2000) When
References member homogeneity is needed in work teams: a meta-
analysis. Small Group Research, 31, 305–327.
Acs, Z.J. and Audretsch, D.B. (1990) Innovation and Small Bradley, S.W., Wiklund, J., and Shepherd, D.A. (2010)
Firms. Boston: MIT Press. Swinging a double-edged sword: the effect of slack on
Acs, Z.J. and Audretsch, D.B. (1991) R&D, firm size and entrepreneurial management and growth. Journal of
innovative activity. In: Acs, Z.J. and Audretsch, D.B. Business Venturing, 26, 537–554.
(eds), Innovation and Technological Change: An Inter- Carton, A.M. and Cummings, J.N. (2012) A theory of sub-
national Comparison. New York: Harvester Wheatsheaf, groups in work teams. Academy of Management Review,
pp. 39–59. 37, 441–470.
Aiken, L.S. and West, S.G. (1991) Multiple Regression: Test- Castellacci, F. (2008) Technological paradigms, regimes
ing and Interpreting Interactions. Newbury Park: Sage. and trajectories: manufacturing and service industries in
Ancona, D. and Caldwell, D. (1992) Demography and a new taxonomy of sectoral patterns of innovation.
design: predictors of new product team performance. Research Policy, 37, 978–994.
Organization Science, 3, 321–241. Chandrasekaran, A. and Linderman, K. (2015) Managing
Atuahene-Gima, K. and Evangelista, F. (2000) Cross-func- knowledge creation in high-tech R&D projects: a multi-
tional influence in new product development: an method study. Decision Sciences, 46, 267–300.

C 2016 RADMA and John Wiley & Sons Ltd

V R&D Management 47, 2, 2017 325
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

Chandy, R.K. and Tellis, G.J. (1998) Organizing for radical Ettlie, J.E. and Rubenstein, A.H. (1993) Firm size and
product innovation: the overlooked role of willingness to product innovation. Journal of Product Innovation Man-
cannibalize. Journal of Marketing Research, 35, 474– agement, 4, 89–108.
487. Evangelista, R. (2000) Sectoral patterns of technological
Chandy, R.K. and Tellis, G.J. (2000) The incumbent’s change in services. Economics of Innovation and New
curse? Incumbency, size and radical product innovation. Technologies, 9, 183–221.
Journal of Marketing, 64, 1–17. Faems, D. and Subramanian, A.M. (2013) R&D manpower
Cohen, W.M. and Levinthal, D.A. (1990) Absorptive and technological performance: the impact of demo-
capacity: a new perspective on learning and innovation. graphic and task-related diversity. Research Policy, 42,
Administrative Science Quarterly, 35, 128–152. 1624–1633.
Cox, T., Lobel, S.A. and Mcleod, P.L. (1991) Effects of Fenwick, G.D. and Neal, D.J. (2001) Effect of gender com-
ethnic group cultural differences on cooperative and position on group performance. Gender Work & Organi-
competitive behavior on a group task. Academy of Man- zation, 8, 205–225.
agement Journal, 34, 827–847. Fernandez Sastre, J. (2015) The impact of R&D teams’ gen-
Dahlin, K.B., Weingart, L.R., and Hinds, P.J. (2005) Team der diversity on innovation outputs. International Journal
diversity and information use. Academy of Management of Entrepreneurship and Small Business, 24, 142–162.
Journal, 48, 1107–1123. Filippucci, C., Drudi, I., and Papalia, R.B. (1996) Testing
Daily, C.M., Dalton, D.R., and Cannella, A.A. (2003) Cor- the relevance of Tobin’s approach for modeling con-
porate governance: decades of dialogue and data. Acad- sumption. Economic Notes, 25, 225–247.
emy of Management Review, 28, 371–382. Frazier, P.A., Tix, A.P., and Barron, K.E. (2004) Testing
Damanpour, F. (1992) Organizational size and innovation. moderator and mediator effects in counseling psychology
Organization Studies, 13, 375–402. research. Journal of Counseling Psychology, 51, 115–134.
Gallouj, F. and Weinstein, O. (1997) Innovation in serv-
Dawson, J.F. (2014) Moderation in management research:
ices. Research Policy, 26, 537–556.
what, why, when, and how. Journal of Business and Psy-
Gibbons, D. and Olk, P.M. (2003) Individual and structural
chology, 29, 1–19.
origins of friendship and social position among profes-
De Faria, P., Lima, F., and Santos, R. (2010) Cooperation
sionals. Journal of Personality and Social Psychology,
in innovation activities: the importance of partners.
84, 340–351.
Research Policy, 39, 1082–1092.
Gibson, C. and Vermeulen, F. (2003) A healthy divide:
Dessler, G. (2001). Leading People and Organisation in
subgroups as a stimulus for team learning behavior.
21st Century. London, England: Prentice Hall.
Administrative Science Quarterly, 48, 202–239.
Dıaz-Garcıa, C., Gonzalez-Moreno, A., and Saez-Martınez,
Gonzalez, J.A. and Denisi, A.S. (2009) Cross-level effects
F.J. (2013) Gender diversity within R&D teams: its
of demography and diversity climate on organizational
impact on radicalness of innovation. Innovation: Man-
attachment and firm effectiveness. Journal of Organiza-
agement, Policy & Practice, 15, 149–160. tional Behavior, 30, 21–40.
Downs, G.W. and Mohr, L.B. (1976) Conceptual issues in Goodstein, J., Gautam, K., and Boeker, W. (1994) The
the study of innovation. Administrative Science Quar- effects of board size and diversity on strategic change.
terly, 21, 700–714. Strategic Management Journal, 15, 241–250.
Dunbar, K. (1995). How scientists really reason: scientific Haas, H. (2010) How can we explain mixed effects of
reasoning in real-world laboratories. In: Sternberg, R.J. diversity on team performance? A review with emphasis
and Davidson, J.E. (eds), The Nature of Insight. Cam- on context. Equality, Diversity and Inclusion. An Inter-
bridge, MA: The MIT Press, pp. 365–395. national Journal, 29, 458–490.
Dunbar, K. (1997). How scientists think: on-line creativity Hagedoorn, J. and Cloodt, M. (2003) Measuring innovative
and conceptual change in science. In: Ward, T.B., Smith, performance: is there an advantage in using multiple
S.M. and Vaid, J. (eds), Creative Thought: An Investiga- indicators? Research Policy, 32, 1365–1379.
tion of Conceptual Structures and Processes. Washington, Hair, J.F., Black, W.C., Babin, B.J., and Anderson, R.E.
DC: American Psychological Association, pp. 461– (2010) Multivariate Data Analysis, 7th edn. Upper Sad-
493. dle River, NJ: Prentice Hall-International.
Dyer, J.H. and Singh, H. (1998) The relational view: coop- Harrison, D.A. and Klein, K.J. (2007) What’s the differ-
erative strategy and sources of interorganizational com- ence? Diversity constructs as separation, variety or dis-
petitive advantage. Academy of Management Review, parity in organisations. Academy of Management
23, 660–679. Review, 32, 1199–1228.
Egan, T.M. (2005) Creativity in the context of team diver- Harrison, D.A., Price, K.H., and Bell, M.P. (1998) Beyond
sity: team leader perspective. Advances in Developing relational demography: time and the effects of surface-
Human Resources, 7, 207–225. and deep-level diversity on work group cohesion. Acad-
Ely, R.J. and Thomas, D.A. (2001) Cultural diversity at emy of Management Journal, 41, 96–107.
work: the effects of diversity perspectives on work group Harrison, D.A., Price, K.H., Gavin, J.H., and Florey, A.T.
processes and outcomes. Administrative Science Quar- (2002) Time, teams, and task performance: changing
terly, 46, 229–273. effects of surface- and deep-level diversity on group

326 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
 Diversity is strategy

functioning. Academy of Management Journal, 45, conflict, and performance in workgroups. Administrative
1029–1045. Science Quarterly, 44, 741–763.
Henkel, J. (2006) Selective revealing in open innovation Joshi, A. and Jackson, S. (2003) Managing workforce
processes: the case of embedded Linux. Research Policy, diversity to enhance cooperation in organizations. In:
35, 953–969. West, M.A., Tjosvold, D., and Smith, K.G. (eds), Inter-
Hitt, M.A., Hoskisson, R.E., Johnson, R.A., and Moesel, national Handbook of Organizational Teamwork and
D.D. (1996) The market for corporate control and firm Cooperative Working. Hoboken, NJ: Wiley, pp. 277–
innovation. The Academy of Management Journal, 39, 296.
1084–1119. Joshi, A. and Roh, H. (2009) The role of context in work
Hoegl, M. (2005) Smaller teams–better teamwork: how to team diversity research: a meta-analytic review. Acad-
keep project teams small. Business Horizons, 48, 209–214. emy of Management Journal, 52, 599–627.
Hong, L. and Page, S.E. (2004) Groups of diverse problem Katz, R. and Tushman, M. (1979) Communication pattners,
solvers can outperform groups of high-ability problem project performance, and task characteristics: an
solvers. Proceedings of the National Academy of Scien- empirical evaluation and integration in an R&D setting.
ces of the United States of America, 101, 16385–16389. Organizational Behaviour and Human Performance, 23,
Horwitz, S. and Horwitz, I. (2007) The effects of team diver- 139–162.
sity on team outcomes: a meta-analytic review of team van Knippenberg, D., De Dreu, C., and Homan, A.C.
demography. Journal of Management, 33, 987–1015. (2004) Work group diversity and group performance: an
Iseke, A., Kocks, B., Schneider, M.R., and Schulze- integrative model and research agenda. Journal of
Bentrop, C. (2015) Cross-cutting organizational and Applied Psychology, 89, 1008–1022.
demographic divides and the performance of research van Knippenberg, D. and Schippers, M.C. (2007) Work
and development teams: two wrongs can make a right. group diversity. Annual Review of Psychology, 58, 515–
R&D Management, 45, 23–40. 541.
Jackson, S.E. (1996) The consequences of diversity in mul-
Langerak, F., Hultink, E.J., and Griffin, A. (2008) Explor-
tidisciplinary work teams. In: West, M.A. (ed.), The Psy-
ing mediating and moderating influences on the links
chology of Groups at Work. Chichester, UK: Wiley, pp.
among cycle time, proficiency in entry timing, and new
53–76.
product profitability. Journal of Product Innovation
Jackson, S.E. and Joshi, A. (2004) Diversity in social con-
Management, 25, 370–385.
text: a multi-attribute, multi-level analysis of team diver-
Lau, D. and Murnighan, J.K. (1998) Demographic diversity
sity and performance. Journal of Organizational
and faultlines: the compositional dynamics of organiza-
Behavior, 25, 675–702.
tional groups. Academy of Management Review, 23,
Jackson, S.E., May, K., and Whitney, K. (1995) Under-
325–340.
standing the dynamics of diversity in decision making
Lau, D.C. and Murnighan, J.K. (2005) Interactions within
teams. In: Guzzo, R. and Salas, E. (eds), Team Effective-
groups and subgroups: the effects of demographic fault-
ness and Decision Making in Organizations. San Fran-
lines. Academy of Management Journal, 48, 645–659.
cisco, CA: Jossey-Bass, pp. 204–261.
Jackson, S.E. and Schuler, R.S. (1995) Understanding Laursen, K. and Salter, A. (2006) Open for innovation: the
human resource management in the context of organiza- roles of openness in explaining innovation performance
tions and their environments. Annual Review of Psychol- among UK manufacturing firms. Strategic Management
ogy, 46, 237–264. Journal, 27, 131–150.
Janis, I.L. (1972) Victims of Groupthink. New York: Lin, J.Y. (2014) Effects on diversity of R&D sources and
Houghton Mifflin. human capital on industrial performance. Technological
Jehn, K.A. and Bezrukova, K. (2010) The faultline activa- Forecasting and Social Change, 85, 168–184.
 Cabello Medina, C., Carmona Lavado,
Lopez Cabrales, A.,
tion process and the effects of activated faultlines on
coalition formation, conflict, and group outcomes. A., and Valle Cabrera, R. (2008) Managing functional
Organizational Behavior and Human Decision Process, diversity, risk taking and incentives for teams to achieve
112, 24–42. radical innovations. R&D Management, 38, 35–50.
Jehn, K.A., Bezrukova, K., and Thatcher, S. (2008) Con- Lundvall, B.A. (1988) Innovation as an interactive process:
flict, diversity, and faultlines in workgroups. In: De from user-producer interaction to the national system of
Dreu, C.K.W. and Gelfand, M. (eds), The Psychology of innovation. In: Dosi, G., Freeman, C., Nelson, R.R., Sil-
Conflict and Conflict Management in Organizations. verberg, G., and Soete, L. (eds), Technical Change and
New York, NY: Erlbaum Psych Press, pp. 179–210. Economic Theory. London: Pinter Publishers, pp. 349–
Jehn, K.A., Chadwick, C., and Thatcher, S.M.B. (1997) To 369.
agree or not to agree: the effects of value congruence, Mairesse, J. and Mohnen, P. (2010) Using innovations sur-
individual demography dissimilarity, and conflict on veys for econometric analysis. In: Hall, B.H. and Rosen-
workgroup outcome. International Journal of Conflict berg, N. (eds), Handbook of the Economics of
Management, 8, 287–305. Innovation. Amsterdam: Elsevier B.V, pp. 1129–1155.
Jehn, K.A., Northcraft, G.B., and Neale, M.A. (1999) Why Malerba, F., Orsenigo, L., and Peretto, P. (1997) Persistence
differences make a difference: a field study of diversity, of innovative activities, sectoral patterns of innovation

C 2016 RADMA and John Wiley & Sons Ltd

V R&D Management 47, 2, 2017 327
Marian Garcia Martinez, Ferdaous Zouaghi and Teresa Garcia Marco

and international technological specialisation. Interna- Page, S.E. (2007) The Difference: How the Power of Diver-
tional Journal of Industrial Organisation, 15, 801–826. sity Creates Better Groups, Firms, Schools and Societies.
Mauleon, E. and Bordons, M. (2014) Indicadores de activi- Princeton, NJ: Princeton University Press.
dad tecnologica por genero en Espa~na a traves del estu- Papalia, R. and Di Iorio, F. (2001) Alternative error term
dio de patentes europeas. Revista Espa~ nola de specifications in the log-Tobit model1. In: Borra, S.,
Documentaci on Cientıfica, 37, e043. Rocci, R., Vichi, M., and Schader, M. (eds), Advances in
Mehra, A., Kilduff, M., and Brass, D.J. (1998) At the mar- Classification and Data Analysis. Berlin Heidelberg:
gins: a distinctiveness approach to the social identity and Springer, pp. 185–192.
social networks of underrepresented groups. The Acad- Pelled, L., Eisenhardt, K., and Xin, K. (1999) Exploring the
emy of Management Journal, 41, 441–452. black box: an analysis of work group diversity, conflict,
Milliken, F.J. and Martins, L.L. (1996) Searching for com- and performance. Administrative Science Quarterly, 44,
mon threads: understanding the multiple effects of diver- 1–28.
Podsakoff, P., Mackenzie, S., Lee, J., and Podsakoff, N.
sity in organizational groups. Academic of Management
(2003) Common method biases in behavioral research: a
Review, 21, 402–433.
critical review of the literature and recommended rem-
Mintzberg, H. (1983) Power in and around Organizations.
edies. Journal of Applied Psychology, 88, 879–903.
Englewood Cliffs, NJ: Prentice-Hall, Inc.
Post, C., De Lia, E., Ditomaso, N., Tirpak, T.M., and
Molleman, E. (2005) Diversity in demographic characteris- Borwankar, R. (2009) Capitalizing on thought diversity for
tics, abilities and personality traits: do faultlines affect innovation. Research-Technology Management, 52, 14–25.
team functioning? Group Decision and Negotiation, 14, Rico, R., Sanchez-Manzanares, M., Antino, M., and Lau,
173–193. D.C. (2012) Bridging team faultlines by combining task
Nemeth, C.J. and Goncalo, J.A. (2005) Influence and per- role assignment and goal structure strategies. Journal of
suasion in small groups. In: Brock, T.C. and Green, Applied Psychology, 97, 407–420.
M.C. (eds), Persuasion: Psychological Insights and Per- Roberge, M. and Van Dick, R. (2010) Recognizing the ben-
spectives. London: Sage Publications. efits of diversity: when and how does diversity increase
Neter, J., Kutner, M.H., Nachtsheim, C.J., and Wasserman, group performance? Human Resource Management
W. (1996) Applied Linear Statistical Models, 4th edn. Review, 20, 295–308.
Chicago: Irwin/McGraw-Hill. Sethi, R., Smith, D.C., and Park, C.W. (2001) Crossfunc-
Nielsen, S. and Huse, M. (2010a) The contribution of women tional product development teams, creativity, and the
on boards of directors: going beyond the surface. Corpo- innovativeness of new consumer products. Journal of
rate Governance: An International Review, 18, 136–148. Marketing Research, 38, 73–85.
Nielsen, S. and Huse, M. (2010b) Women directors’ contri- Shore, L.M., Chung-Herrera, B.G., Dean, M.A., Ehrhart,
bution to board decision-making and strategic involve- K.H., Jung, D.I., Randel, A.E., and Singh, G. (2009)
ment: the role of equality perception. European Diversity in organizations: where are we now and where
Management Review, 7, 16–29. are we going? Human Resource Management Review,
Nishii, L.H. and Goncalo, J.A. (2008) Demographic fault 19, 117–133.
lines and creativity in diverse groups. In: Phillips, K.W., Simons, T., Pelled, L.H., and Smith, K.A. (1999) Making
Mannix, E.A., and Neale, M.A. (eds), Research on Man- use of difference: diversity, debate, and decision compre-
aging Groups and Teams. Greenwich, CT: JAI Press, pp. hensiveness in top management teams. Academy of Man-
1–26. agement Journal, 42, 662–673.
Singh, V., Terjesen, S., and Vinnicombe, S. (2008) Newly
Nonaka, I. and Takeuchi, H. (1995) The Knowledge-
appointed directors in the boardroom: how do women and
Creating Company. New York, Oxford: Oxford Univer-
men differ? European Management Journal, 26, 48–58.
sity Press.
Sirilli, G. and Evangelista, R. (1998) Technological innova-
O’reilly, C.A., Williams, K.Y., and Barsade, S. (1998)
tion in services and manufacturing: results from Italian
Group demography and innovation: does diversity help?
surveys. Research Policy, 27, 881–899.
In: Gruenfield, D. and Neale, M.A. (eds), Research on
Sofka, W. and Grimpe, C. (2010) Specialised search and
Managing in Groups and Teams. Stamford, CT: JAI innovation performance: evidence across Europe. R&D
Press, pp. 183–207. Management, 40, 310–323.
OECD (2002) Frascati Manual. Proposed Standard Prac- Sollner, R. (2010). Human capital diversity and product
tice for Surveys on Research and Experimental Develop- innovation: a micro-level analysis. Jena Economic
ment. Paris: OECD Publications. Research Papers- 027.
OECD (2005) Oslo Manual. Guidelines for Collecting and Spickermann, A., Zimmermann, M., and Von der Gracht,
Interpreting Innovation, 3rd edn. Paris: OECD H.A. (2014) Surface- and deep-level diversity in panel
Publications. selection — exploring diversity effects on response
Østergaard, C.R., Timmermans, B., and Kristinsson, K. behaviour in foresight. Technological Forecasting and
(2011) Does a different view create something new? The Social Change, 85, 105–120.
effect of employee diversity on innovation. Research Sutton, R.I. (2007) Weird Ideas That Work: How to Build a
Policy, 40, 500–509. Creative Company. New York: Free Press.

328 R&D Management 47, 2, 2017 C 2016 RADMA and John Wiley & Sons Ltd
V
 Diversity is strategy

Swanson, R. and Holton, E. (2009) Foundations of Human Wood, W. (1987) Meta-analytic review of sex differences in
Resource Development, 2nd edn. San Francisco, CA: group performance. Psychological Bulletin, 102, 53–71.
Berrett-Koehler. Woodman, R.W., Sawyer, J.E., and Griffin, R.W. (1993)
Tajfel, H. (1978) Differentiation between Social Groups: Toward a theory of organizational creativity. The Acad-
Studies in the Social Psychology of Intergroup Relations. emy of Management Review, 18, 293–321.
London: Academic Press.
Yang, Y., Narayanan, V.K., and Zahra, S. (2009) Develop-
Talke, K., Salomo, S., and Rost, K. (2010) How top man-
ing the selection and valuation capabilities through
agement team diversity affects innovativeness and per-
learning: the case of corporate venture capital. Journal
formance via the strategic choice to focus on innovation
of Business Venturing, 24, 261–273.
fields. Research Policy, 39, 907–918.
Thatcher, S.M.B. and Patel, P.C. (2011) Demographic Zenger, T.R. and Lawrence, B.S. (1989) The differential
faultlines: a meta-analysis and synthesis of the literature. effects of age and tenure distributions on technical
Journal of Applied Psychology, 96, 1119–1139. communication. Academy of Management Journal, 32,
Thatcher, S.M.B. and Patel, P.C. (2012) Group faultlines: a 353–376.
review, integration, and guide to future research. Journal
of Management, 38, 969–1009.
Tsui, A.S., Ashford, S.J., Clair, L.S., and Xin, K.R. (1995) Note
Dealing with discrepant expectations: response strategies
and managerial effectiveness. Academy of Management 1. R&D employment data are only available for internal
Journal, 38, 1515–1543. R&D performers.
Turner, J.C., Hogg, M.A., Oakes, P.J., Reicher, S.D., and
Wetherell, M.S. (1987) Rediscovering the Social Group: Marian Garcia Martinez is a Professor of Market-
A Self-Categorization Theory. New York: Blackwell. ing and Innovation at Kent Business School, UK.
Van der Vegt, G. and Janssen, O. (2003) Joint impact of Her primary area of research is marketing-led inno-
interdependence and group diversity on innovation. vation, with a particular focus on new product mar-
Journal of Management, 29, 729–751.
keting, and open innovation and value co-creation
Williams, H., Parker, S., and Turner, N. (2007) Perceived
models.
dissimilarity and perspective taking within work teams.
Group and Organization Management, 32, 569–597. Ferdaous Zouaghi is a Research Assistant at the
Williams, K. and O’reilly, C. (1998) Demography and
Public University of Navarra, Spain. Her PhD
diversity in organizations: a review of 40 years of
research. In: Staw, B. and Cummings, L. (eds), Research research focuses on innovation activities in the Agri-
in Organizational Behavior. Greenwich: JAI Press, pp. food Industry.
77–140.
Teresa Garcia Marco is Associated Professor in
Wittenbaum, G.M. and Stasser, G. (1996) Management of
information in small groups. In: Nye, J.L. and Brower, Business Organisation at the Public University of
A.M. (eds), What’s Social about Social Cognition? Navarra, Spain. Her research interests include antece-
Research on Socially Shared Cognition in Small Groups. dents of innovation performances and open innova-
Thousand Oaks, CA: Sage, pp. 3–28. tion strategies.

C 2016 RADMA and John Wiley & Sons Ltd

V R&D Management 47, 2, 2017 329
Copyright of R&D Management is the property of Wiley-Blackwell and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.