Working With More Creative Coworkers and New Product Development Team Performance

At the Individual Level

Our arguments are based on the perspective of supplementary fit in the person-group fit theory (Cable & Edwards, 2004), which proposes that similarity in a person’s attribute to the attributes of the other individuals on their team may lead to positive outcomes (Kristof-Brown et al., 2005). Because members are similar, workers in teams in which members possess homogenous attributes tend to like one another, can work smoothly together, and possess similar expectations about their colleagues on the team (Cable & Edwards, 2004; Muchinsky & Monahan, 1987). Drawing on previous research on social exchange relationships in work teams (Bernerth et al., 2008), we propose that if an individual possesses a level of openness to experience personality similar to that of other members of the team, this person will experience a higher level of team member exchange (TMX).

TMX is a form of social exchange relationship a worker has with other members of the team and is characterized by long-term and unspecified mutual obligations (Seers et al., 1995). The personality literature finds that highly open persons think outside the box and see things from new perspectives (Costa & McCrae, 1992). Openness personality is also found to be positively related to creativity, intelligence, and social problem-solving skills (Buss, 1991). According to the theory of reciprocity, a long-term, beneficial relationship can only be maintained with a mutual and balanced exchange (Gouldner, 1960). In an NP development team, members need to come up with creative solutions to deal with problems in the NP developmental process to deliver successful NPs (Amabile, 1988; Woodman et al., 1993). When a highly open engineer works with a group of colleagues who possess similar levels of high openness personality, team members may challenge one another mentally and benefit from their colleagues’ propositions in the NP developmental process. This mutually beneficial exchange of ideas nurtures a work environment of the reciprocal contribution of proposals and feedback, an essential precondition for a high TMX relationship (Seers, 1989).

Conversely, if a highly open engineer works with a group of intellectually dull and close-minded teammates, it is likely that this highly open person may feel their suggestions cannot be met with commensurate responses. This condition cannot nurture a work environment of the reciprocal contribution of ideas, which is required for a high TMX. The same condition occurs when a close-minded person works with a group of highly open colleagues. When other highly open teammates come up with creative suggestions and find that the focal person cannot respond with comparably inspiring feedback, it is unlikely that the focal person may develop a mutually beneficial exchange relationship with other highly open colleagues.

Though few studies have examined similarity in personality and team-member exchange (TMX), research into the social exchange relationship with the supervisor provides indirect support for our argument. For instance, Bernerth et al. (2008) collect data from the United States and find that similarity in openness personality between workers and supervisors, which the authors designate personality congruence, is positively related to the leader-member exchange (LMX) relationship. The above finding suggests, in an analogous manner, that when a person works with colleagues possessing a large difference in openness personality, it is unlikely that the person may develop a high exchange relationship with their teammates. Thus, assembling workers with similar levels of openness personality should be positively related to an individual’s experience of social exchange relationships with teammates. This study therefore proposes:

TMX and Creativity

Previous research finds that TMX can enhance workers’ self-efficacy, which is defined as a person’s belief in their capability of attaining given outcomes (Bandura, 1997). The rationale supporting the TMX-Self-efficacy relationship rests on the social cognitive theory (Bandura, 1982), which argues that a person’s belief about their self-capability may be influenced by their interactions with similar social models (Bandura, 1997; Brower & Nurius, 1993). Because individuals on the same team share common goals and work on related tasks (Seers, 1989), workers may choose teammates to observe as social models (Ford & Seers, 2006). High-quality TMX may allow a person to interact with teammates sufficiently and observe their behaviors at work to obtain effective skills for managing job demands (Tse & Dasborough, 2008). Thus, a team member may obtain a sense of personal efficacy from interacting with other team members (Bandura, 1982).

Bandura (2001) further argues that persons with high self-efficacy are more likely to set challenging goals and exert greater effort to attain these objectives. Scholars of creativity thus draw on this rationale to propose that workers with high self-efficacy will proactively initiate creative solutions on their jobs (Tierney & Farmer, 2002, 2004). Examining Chinese workers, Liao et al. (2010) further demonstrate that TMX can strengthen workers’ self-efficacy and stimulate their creativity. The above findings confirm that TMX can ultimately help members think of creative ideas. Other research also suggests that TMX may stimulate individual creativity. For example, Kirrane et al. (2019) find that TMX is positively related to creativity. Ghosh et al. (2019) find that TMX is positively related to individual innovation work behavior, which refers to whether a person has implemented creative ideas into innovative behavior. Moreover, Ehrhardt and Sharif (2019) find that a coworker relationship increases an individual’s career self-efficacy and leads one to explore new career opportunities. The above findings suggest that high-quality TMX is conducive to the generation of creative ideas.

In developing H1, we have proposed that team members’ similarity in openness personality with teammates can enhance TMX. In this paragraph, we further argue that TMX is an antecedent of creativity (Liao et al., 2010). Therefore, we propose at the individual level that gathering team members with similar levels of openness personality may strengthen team members’ TMX to help individuals think of creative ideas:

Theories of Individual Creativity and Team Innovation

Several theories of creativity and innovation have proposed that the creative outcome of a group is a function of the aggregate of group members’ creativity (Amabile, 1988; Woodman et al., 1993). These theories argue that team members’ creativity in aggregate may help the NPD team perform better to improve NP performance (Anderson et al., 2014; Woodman et al., 1993). Following the above rationale, we aggregate the effects at the individual level into the team level. As we have already proposed in H2 that bringing together members with similar levels of openness personality can stimulate workers’ creativity via enhanced TMX, we propose that team members’ aggregated similarity in openness personality with teammates may nurture team members’ TMX and improve the performance of the NPD team, that is, improve NP performance, defined as the extent to which the new product developed attains its goals in terms of sales, market share, profit, and overall performance (Sethi & Iqbal, 2008).

Although few studies have examined the effect of team members’ aggregated similarity in openness personality with NP performance, research provides indirect evidence that suggests working with teammates of less diversity in personality is good for team performance. For instance, Liang et al. (2015) examine workers in Taiwan and find that team diversity in conscientiousness, agreeableness, and neuroticism personality are negatively related to team cohesion and cooperation and deter team helping behavior. Since helping behaviors may improve team performance (Podsakoff et al., 2000), the findings of Liang et al. (2015) suggest that diversity in team personality may not always help team performance, while similarity in team personality may improve team outcomes. Examining information system development teams, Trimmer et al. (2002) find team diversity in emotional stability and conscientious personality is positively related to relational conflict. Since relational conflict is found to negatively influence team performance (De Dreu & Weingart, 2003), Trimmer et al. (2002) provide indirect evidence to suggest that team aggregated similarity in openness personality may help the performance of NPD teams. Therefore we propose:

The Moderating Effect

Building on previous theories of creativity and innovation (Anderson et al., 2014; Woodman et al., 1993), Perry-Smith and Mannucci (2017) propose a four phases model concerning how creative ideas are transformed into successful innovation of work teams. The first phase is creative idea generation, which concerns the generation of new ideas; the second phase is new idea elaboration, which concerns the person who initiated the creative suggestion further clarifying and developing his or her proposal; the third phase is idea championing, which concerns promoting the new idea proposed to obtain a green light to develop the new idea; the fourth phase is idea implementation, which concerns implementing the new idea into a tangible new product. Perry-Smith and Mannucci (2017) argue that especially in the idea championing and the idea implementation phases, individuals who come up with creative ideas need cooperation and support from their teammates to implement their novel proposals into successful new products.

The personality literature finds that highly open persons like to experience new things, are more acceptable to unconventional ideas, and are more willing to try new activities (McCrae & John, 1992). Because highly open individuals can make novel connections among remotely related ideas, highly open persons are more creative (Shalley et al., 2004). People ranking low in openness personality find comfort in familiar surroundings and may surrender the opportunities to try new experiences. They stick to tried and true practices and seek refuge in familiar environments (Costa & McCrae, 1992).

We derive hypotheses by integrating the idea-journey theory of creativity and innovation (Perry-Smith & Mannucci, 2017) with findings in the openness personality literature and propose that a high aggregated team openness personality can facilitate creativity generation, and creativity championing and implementation phases in the team innovation process to help an innovation team attain good NP performance. Highly open persons are better at thinking of creative ideas (Shalley et al., 2004), which can facilitate the idea generation phase (first step) when innovation teams develop new products. Because highly open persons are also more accepting of unconventional ideas and are more willing to try out new alternatives (McCrae & John, 1992), members in NPD teams with a high aggregated openness personality are more likely to accept and support the creative proposal initiated by team members instead of bogging down these creative propositions in the team innovation process. This facilitates the idea championing and idea implementation phases and help the innovation team deliver successful new products. Therefore, teams with a high aggregated open-personality do not need to gather members with similar levels of openness personality to stimulate TMX and come up with creativity. The positive effect of team aggregated similarity in openness personality with teammates on NP performance will thus be less salient in teams with a high aggregated openness personality.

If the NPD team manager cannot form a team with members all high in open personality, as long as the manager gathers engineers with similar levels of open personality, members can still think of creative ideas via better TMX (H2). According to the idea journey model of creativity and innovation (Perry-Smith & Mannucci, 2017), because gathering members with a similar, although not necessarily high, level of openness personality may nurture TMX to stimulate creativity, team aggregated similarity in members’ openness personality can facilitate the idea generation phase of the team innovation process to help the team deliver successful new products.

Moreover, the TMX literature finds that workers with high TMX possess mutual trust and receive reciprocal caring from teammates (Dayan et al., 2009). They identify more and possess higher levels of trust toward coworkers (Farmer et al., 2015), and may offer help to one another because they believe their good deeds will be reciprocated (Molm et al., 2000). According to the idea journey model of creativity and innovation (Perry-Smith & Mannucci, 2017), assembling workers with similar, although not necessarily high, levels of open personality may nurture TMX among members so that members are more likely to support the creative ideas proposed by their colleagues instead of bogging down these innovative propositions. This facilitates the idea championing and idea implementation phases of the team innovation process and help the NPD team implement new ideas into successful new products.

Therefore, if NPD team managers cannot recruit enough workers with high levels of open personality to join their team, an alternative approach to improving NP performance is to gather members of similar, although not necessarily high, levels of openness personality. This is especially important for teams with a low aggregated open-personality, since members with similar levels of openness personality may nurture TMX and stimulate workers’ creative solutions. Moreover, since gathering members with similar levels of openness personality can nurture TMX among members, doing so breeds members’ support for creative propositions and facilitates the idea championing and implementation phases to help the team implement creative ideas into successful new products. Thus, the positive effect of team aggregated similarity in openness personality with teammates on NP performance will be especially salient in teams with a low aggregated openness personality.

Finally, sometimes the NPD team is composed of a small percentage of members high in open personality while most other members are low in openness personality. Although highly open persons may stimulate creative thoughts in the idea generation phase (Schilpzand et al., 2011), differences in levels of open personality adversely affect TMX (H2), which hinders cooperation and support among members in the idea championing and idea implementation phases. According to the idea-journey theory of creativity and innovation (Perry-Smith & Mannucci, 2017), such NPD teams do not stand a good chance of delivering successful new products.

Summarizing the above arguments, we see that the effect of gathering members with similar levels of open personality on NP performance will be especially salient when team members in aggregate possess low levels of openness personality. We propose:

Samples

The survey was conducted by consulting R&D managers of optoelectronic manufacturers in Taiwan. The purpose of this survey and the data needed were explained to the questionnaire takers in the NP development teams, and one NPD team from each company was assigned to this task. With a response rate of about 24%, 61 companies participated in this survey. The first wave of questionnaires was sent to the members of the teams assigned based on openness personality and the control variables, as explained in the following paragraphs. Since the information of the team members must be aggregated to form team level data, 90% of the information collected successfully from the team members was used. In this way, the final effective responses comprised 274 R&D engineers working in NPD teams among 56 firms in the first wave of data collection.

Three months after completion of the first wave of data collection, the second set of questionnaires were sent to these NPD engineers for assessing their team-member exchange relationship. After collecting responses from these NPD engineers successfully, which took approximately 2 months, another wave of questionnaires was sent to the leaders of these NPD teams for evaluation of the creativity performance of team members and NP’s performance. Three teams were dissolved during the data collection, hence the data collected were excluded from this study.

In total, 263 matched responses from 53 NPD teams among 53 companies were collected. Following previous studies, the responses were divided into two parts: the earliest 75% received and the latest 25% (Guthrie, 2001). T-tests found no significant difference between the early and late-respondents in terms of all the major research variables. 40.7% of the R&D engineers participants were female and 59.3% were male. Among the supervisors who evaluated the participants’ creativity, 30.9% were female and 69.1% were male. In the 53 participating teams, 14 teams had 4 members, 30 teams had 5 members, 6 teams had 6 members, and 3 teams had 7 members.

Measures

We surveyed all constructs with scales used in previous studies. Since these scales were developed in English, we followed previous authors to translate them into Chinese with back translation to ensure the accuracy of the translation (Brislin, 1986). To ensure the wordings of the survey questions possess adequate meaning to the participants and deploying the study context properly, eight NPD team leaders from these firms were also interviewed. All constructs were rated on a 7-point Likert scale from strongly agree (7) to strongly disagree (1) unless specified otherwise.

Following Liao et al. (2008), we used the ten-item scale in the International Personality Item Pool (IPIP) to measure openness personality (Goldberg, 1999). A sample item is: “I have a vivid imagination.” Following Bernerth et al. (2008), this study measured similarity in openness personality with teammates using the square root of the sum of the squared differences (i.e., Euclidian distance, D) between the openness personality rated by the focal person and that rated by other teammates. Thus, for a team with N members, there will be N − 1 pairs of differences in openness personality. Following Tsui and O’reilly (1989), we divided the D score by N − 1, where N equals the number of persons on the team. The result was also coded in the negative form to represent the level of similarity in the respondent’s openness personality with that of other teammates. A larger score (D/N − 1) represents a large difference in openness personality between the focal person and all other teammates on the team. This study then adopts 1-difference to represent similarity.

We asked the respondents to answer the 10-item scale developed by Seers et al. (1995) to assess the quality of their exchange relationship with other team members and followed Liao et al. (2010) to aggregate answers on these 10 items to form one construct called TMX. Sample items are: “I am willing to help finish work that had been assigned to other members on my team,” and “Other members on my team are willing to help finish work assigned to me.” Creativity was measured with Farmer et al.’s (2003) four-item scale. A sample item is: “This employee will try new ideas or methods first.”

According to the idea-journey theory of creativity and innovation (van Knippenberg & Hoever, 2017), the outcome of the creativity and innovation process is the new product delivered by the innovation team. We thus used Sethi and Iqbal’s (2008) four-item scale developed by to measure whether the NP developed by the NPD team attained its goal in terms of sales, market share, profit, and overall performance. Previous authors have used various ways to operationalize the performance of NPD teams, such as using archival data or conducting experiments (Fuchs & Schreier, 2011; Yli-Renko & Janakiraman, 2008). Nevertheless, research finds a strong convergence in objective and subjective measures of NP performance, showing that using subjective measurements is as valid as using objective measures (Baker & Sinkula, 2007; Boso et al., 2013).

Control Variables

To rule out alternative explanations for the findings, we involved four control variables at the individual level. Following Liao et al. (2010), we controlled for the respondents’ demographic characteristics of age and education because previous authors found these demographic variables could affect a person’s relationship with peers (Seers, 1989). Respondents were asked to indicate the final degree they received (1 = high school, 2 = bachelor, 3 = master, doctoral, and post-doctoral research), their age (1 = 25 under, 2 = 36–35, 3 = 36-45, 4 = above 46), and their tenure (1 = under 6 years, 2 = 6–15 years, 3 = 16–20 years, 4 = above 20 years). Moreover, since workers more confident in their abilities are more proficient in thinking of creative ideas in the first place (Liao et al., 2010; Tierney & Farmer, 2002, 2004), this study followed Liao et al. (2010) and controlled for the respondents’ self-efficacy at the beginning of the study (time 1). The respondents’ self-efficacy at time 1 was measured with the eight-item scale developed by Chen et al. (2001); a sample item is: “I believe I can succeed at almost any endeavor to which I set my mind.” Finally, since the panel consultants thought that workers who have more experience in developing new products are more proficient and more confident in their capabilities, we controlled for the respondents’ years of experience in the industry. At the team level, this study controlled for the team’s size, measured as the number of workers on the team because larger teams have more resources to accomplish their assignment and may thus perform better.

Validity

We set openness personality as a one-factor construct composed of 10 items, TMX as composed of 10 items, creativity is composed of 4 items, and self-efficacy as composed of 8 items. The CFA results show that all relevant questionnaire items load significantly on their intended constructs (p < .01) and that the data fit the proposed four-factor model well (χ²/df = 1.64, NFI = 0.88, TLI = 0.94, CFI = 0.95, RMSEA = 0.05). This study also compared alternative models with three factors (χ²/df = 2.68, NFI = 0.80, TLI = 0.85, CFI = 0.87, RMSEA = 0.08, combining self-efficacy and open to experience personality), two factors (χ²/df = 3.59, NFI = 0.74, TLI = 0.77, CFI = 0.79, RMSEA = 0.10, combining self-efficacy, open to experience personality, and TMX), and one factor (χ²/df = 4.07, NFI = 0.70, TLI = 0.72, CFI = 0.75, RMSEA = 0.11, combining self-efficacy, open to experience personality, TMX, and creativity), but found the four-factor model exhibited the best fit with the data.

Composite reliability for all constructs came out satisfactorily, confirming adequate measurement reliability (Table 1 for detailed information). This study also listed the Cronbach’s α values of all constructs as additional support for measurement reliability. Since we found the average variance extracted for all constructs exceeds 0.50 and that all scale items loaded significantly on their hypothesized constructs (t-test, p < .00), these results demonstrate adequate convergent validity (Anderson & Gerbing, 1988; Anderson et al., 1998).

OPEN IN VIEWER

Table 1.

Validity, Mean, Standard Deviation, and Correlation Matrix.

Variable	CR	AVE	α	M	SD	Effic	Open	TMX	Creativity
1. Self-efficacy	0.89	0.51	.90	5.31	0.92	0.71
2. Open to experience personality	0.89	0.45	.90	5.14	0.80	0.39**	0.67
3. Team member exchange	0.92	0.55	.92	5.48	0.71	0.42**	0.52**	0.74
4. Creativity	0.89	0.67	.89	5.19	0.93	0.38**	0.55**	0.50**	0.82

Note. The root square of AVEs is shown on the diagonal line in boldface.

*

p < .05; **p < .01 (two tailed).

To examine discriminate validity, we followed Fornell and Larcker (1981) and compared the root square of AVEs of all relevant constructs (openness personality, TMX, creativity, and self-efficacy) with the correlation of all possible pairs of constructs. The results are listed in Table 1. Since the square roots of the AVEs were all found to be greater than the correlation of all pairs of constructs, this study found support for the discriminant validity of the constructs used in this study.

Theoretical Significance

The creativity and innovation theory argues that the creative outcome of a group is a function of group members’ personality, group composition, group processes, and contextual factors (Woodman et al., 1993). Further, research into creativity finds individuals with a high open personality are more likely to think of creative ideas (Feist, 1998). These findings suggest that NP managers should recruit highly creative workers into NP development teams. When members on the teams all possess high levels of openness personality, that is, with high creative potential, members may think of creative solutions in the NP developmental process to improve the NP’s performance.

Scholars since then have examined how team diversity in personality composition may be related to the creative performance of work teams (Reiter-Palmon, 2017). For instance, Moynihan and Peterson (2004) argue for the cognitive stimulation effect of exposure to highly creative coworkers so that gathering workers with different levels of personality will be good for team creativity performance. Examining graduate student teams, Schilpzand et al. (2011) find that team members’ aggregated openness personality is positively related to team creativity and that teams with the greatest diversity in openness personality perform most creatively. Conversely, Bond and Shiu (1997) find low variability in team openness personality reduces team shared exchange, which may in turn adversely influence team performance. These findings suggest that, even for a personality trait that is found to be helpful to creativity, such as the openness personality, NPD team managers should not recruit all workers high in this trait but should instead gather members with different levels of this trait to improve individual creativity and in turn, enhance NP performance.

This study complements the literature by using the social exchange perspective, a different angle. Drawing on the supplementary fit perspective in the person-group fit theory (Cable & Edwards, 2004), we propose and find at the individual level that by putting together workers with similar levels of open personality, managers can foster workers’ TMX and in turn stimulate individual creativity. Examining data collected from Taiwanese new product development teams, we also find at the team level that team members’ aggregated similarity in openness personality with teammates may improve the performance of the NP developed by the team. By emphasizing the social exchange effect stemming from working with colleagues possessing similar levels of openness personality, an indicator of an individual’s creative potential, we show why and how working with colleagues possessing similar levels of openness personality can stimulate individual creativity.

At the same time, our findings also suggest that gathering team members with different levels of openness personality is not conducive to individual creativity. This unfavorable effect was not yet evident before our study. Our findings also differ from suggestions that may be derived from the literature as mentioned in the previous paragraph. To our knowledge, the social exchange relationship with coworkers is a theoretical perspective scholars have yet to employ in examining how working with more creative colleagues may influence creativity. Thus, the findings of this study may add to scholarly knowledge about how team personality composition may relate to individual creativity.

Our study also complements scholarly knowledge about how team personality composition may be related to the performance of NPD teams. Before this study, scholars posited the cognitive-stimulation effect and proposed that gathering members with different levels of creativity-conducive personality can help team creativity performance (Moynihan & Peterson, 2004; Schilpzand et al., 2011). Combining this argument with the theories of creativity and innovation, which maintain that the creative outcome of a group is a function of the aggregate of group members’ creativity (Amabile, 1988; Woodman et al., 1993), it appears that gathering members with different levels of open personality, an indicator of an individuals’ creative potential, can help an innovation team improve the performance of the new product it develops.

Examining from a social-behavioral perspective, we argue and demonstrate that bringing together workers with similar, although not necessarily high, levels of creative potential, that is, possessing similar levels of openness personality, can increase workers’ social exchange relationships with colleagues to help the team improve NP performance. Research into personality finds that openness personality is normally distributed, which suggests that only a relatively small proportion of persons in our society possess very high levels of openness personality (McCrae & John, 1992). Our findings remind scholars and managers that they do not have to recruit members high in openness personality for every slot in a team. Rather, forming teams of members with similar, although not necessarily high, levels of open personality can be an effective way to generate better performance from the NPD team.

Our study further finds that the beneficial effect of putting together workers with similar levels of openness personality on NP performance only exists when team members in aggregate possess a low, rather than a high, level of the openness personality. This finding suggests that forming teams with members at similar levels of open personality not only can supplement workers’ insufficiency in creative potential but can also create members’ support for creative propositions in the team innovation process to help the innovation team enhance NP performance. Our finding differs from the implications that may be derived from the literature, which predicts that assembling members with different levels of open personality can improve the creativity performance of the team (Moynihan & Peterson, 2004; Schilpzand et al., 2011). This study, using the social-behavioral perspective, a theoretical perspective not yet systematically examined in the relevant literature, demonstrates how team composition of openness personality may be related to the performance of the innovation team. As such, this study complements the literature about teams (Brower, 1989; 1996; Brower & Rose, 1990, 2015).

Practical Implications

Our findings tell managers that in case they cannot find enough workers high in openness personality and thus, in creative potential, they may assemble teams composed of members with similar levels of openness personality. Doing so may create a team setting favorable to TMX and in turn, stimulate creativity from individual workers. Our study demonstrates that assembling teams with members possessing similar, although not necessarily high, levels of openness personality can offset worker deficiencies in creative potential, that is, without high levels of openness personality, to stimulate creative ideas.

At the team level, this study demonstrates that forming teams with members with similar levels of openness personality can facilitate the team innovation process to improve NP performance and that this effect is especially salient when team members in aggregate possess a low, rather than a high, level of openness personality.

Limitations

Inevitably, this study suffers from several limitations, but at the same time also offers opportunities for future research. First, to observe a range of NPD teams, this study collected data on performance with an assessment from team managers. As explained previously, we use managers’ assessment because of the lack of available objective measurements on the performance of the new products the teams developed. Although previous authors find managers’ subjective assessment of NP performance is strongly correlated with objective measurements (Baker & Sinkula, 2007; Boso et al., 2013), other studies find that self-reported measures of creativity are not correlated with the objective measurements (Reiter-Palmon et al., 2012). Sometimes, managers may have better interactions or possess a more positive impression of a certain work team so that their subjective assessment may reflect their general impression of a team instead of its actual performance. Research in the future may collect objective data on NP performance. Results from such a study may enhance the internal validity of the findings from our survey.

However, we believe these concerns may not completely refute the validity of the findings from this study. Since team managers should be quite familiar with the creative performance of their team members and with the performance of the NP that the team developed, surveying data from managers should represent the real situation to a certain extent. Another reason for our use of managers’ evaluation to gauge NP performance is the lack of archival data measuring new products’ performance across various firms. To enhance the external validity of our findings, this study collected information from 53 optoelectronic manufacturing firms in Taiwan. Each of these firms aims at different markets and specializes in different products. Because this study cannot find objective measurements for all these different products, this study chooses to use managers’ assessment because the managers should possess a clear understanding of how the new product their firm developed perform. Since previous research finds a strong correlation between subjective and objective measures of NP performance (Baker & Sinkula, 2007; Boso et al., 2013), this study follows the conventions of the new product development literature (Im & Workman, 2004; Sethi & Iqbal, 2008) and uses managers’ evaluation to assess NP performance. Thus, the findings of this study may still possess a high degree of validity.

Second, our study only examines how similarity in openness personality with teammates may enhance workers’ TMX to improve creativity. Yet, research finds that TMX may have various beneficial effects on team outcomes, such as increasing members’ team commitment, knowledge sharing, and motivation (Jordan et al., 2002; Liu et al., 2011). As this study concentrates on examining creativity without investigating other aspects of performance, future scholars may examine from perspectives other than creativity to enrich scholarly understanding about how and through what mechanisms team personality composition may influence team performance.

Third, our findings cannot be generalized to conclude that similarity in any type of personality may enhance team performance for all kinds of tasks. We argued that similarity in openness personality with teammates may enhance workers’ TMX to improve individual creativity and NPD team performance. An underlying assumption of this argument is that reduced TMX may lower cohesion to adversely affect team performance (Mach et al., 2010). However, personality traits such as conscientiousness may be related to other skills required for the successful accomplishment of team assignments and can thus compensate for a reduced level of team cohesion (Bell, 2007). Later authors may examine other types of personality and other aspects of team performance to complement the relevant literature.

Fourth, although we have collected data with time-lag and information on the independent and dependent variables from different sources to improve the internal validity of our findings, our study cannot validate causality. For instance, previous authors find that high group performance may increase group cohesiveness (Mullen & Copper, 1994). In this way, the mechanism implied in our study, reduced TMX, which may lower cohesion to adversely affect team performance, may be influenced by previous team performance. Because this study examines teams that develop new products, it is difficult to collect information on the team’s previous performance because their very mission is to develop new products. Researchers may conduct experiments to examine the phenomena we have explored to verify the causality of our findings.

A final limitation is that managers may not always find it easy to use the findings from our study in helping them select NPD team members. Recent authors have discussed the problem of applying theories in the field of social psychology and argue that psychological theories are not highly relevant to people’s everyday lives (Berkman & Wilson, 2021). Engineers in Taiwanese manufacturing firms may expect to be selected for new product development jobs based on their professional skills and knowledge, competence, and previous relevant work experience, rather than the level of their open personality. In this way, the findings from our study only indicate a criterion that NPD team managers may consider in helping them make selection decisions when assembling innovation teams. Managers may not always find it easy to apply these criteria in recruiting NPD team members under all circumstances. Moreover, there are other characteristics relevant to creative outcomes aside from open personality, such as motivational styles and creativity styles. Future research may explore the effects of these characteristics on team outcomes to enrich the literature.