Leadership and Creativity

Research Setting, Sample and Procedure

Our survey is conducted in eight Chinese companies that operate in various industries. We first interviewed senior managers from these companies to acquire the permission and confirm that creativity is a characteristic the companies are aiming to achieve. Next, we randomly chose three to ten teams per firm (M = 5.4) and then sent questionnaires to team members (including demographics and independent variable measures) and team leaders (including team and employee creativity measures) through emails. We ultimately received 237 valid questionnaires from team members (response rate of 84.0%) and 43 from team leaders (response rate of 86.0%). Of the participants, 57.0% were men, and 43.0% were women. Their average age was 30.3 years, and the average tenure in their jobs was 8.2 years. The most frequently indicated education level was a bachelor's degree (57.0%), and most participants were technical workers (81.0%).

Measures

We used back-translation to translate our English questionnaire into Chinese. Unless noted otherwise, items were assessed on five-point Likert scale ranging from "1 = strongly disagree" to "5 = strongly agree".

Entrepreneurial Leadership (α = 0.91)

We followed the definition of entrepreneurial leadership in Renko et al. in the present study to show that entrepreneurial leaders direct and encourage employees' behaviours to pursue entrepreneurial goals. Thus, we measured entrepreneurial leadership using eight items from Renko et al.'s ENTRELEAD-scale, which measures employees' perception of their leaders' entrepreneurial leadership qualities. Showing a high level of internal consistency and validity, this scale featuring the role of entrepreneurial leadership style in leading employees has been validated in the organizational behaviour research. One of the sample items is "My supervisor often comes up with ideas of completely new products/services that we could sell". Employees rated each item regarding their supervisor's entrepreneurial leadership.

Creative Self-efficacy (α = 0.76)

We used a three-item measure of creative self-efficacy from Tierney and Farmer to assess employees' creative self-efficacy (e.g. "I have confidence in my ability to solve problems creatively"). Employees were asked to indicate the degree to which the statements accurately describe their efficacy with regard to creative work.

Team Creative Efficacy (α = 0.85)

We used four items to measure team creative efficacy from Tierney and Farmer and Shin and Eom by modifying the creative self-efficacy items to focus on teams' creative efficacy (e.g. "Members of my team have confidence in their abilities to solve problems creatively"). As members constitute the whole team, these items are designed to tap individuals' perceptions of the extent to which each statement describes their team members' shared beliefs in their team's capabilities to perform creative tasks. This team-level efficacy scale appropriately points to the teams' creative capabilities and has been extensively validated in prior research.

Employee Creativity (α = 0.84)

We used four items from Farmer et al. and asked supervisors to report the creativity of each of their employees (e.g. "This employee seeks new ideas and ways to solve problems"). This scale has been developed for the Chinese context to reflect the Chinese view of employee creativity.

Team Creativity (α = 0.81)

Using Shin and Zhou's four-item scale, we asked supervisors to measure their team's creative performance (e.g. "How well does your team produce new ideas?"), with a range from 1 = poor to 5 = excellent.

Control Variables

At the individual level, we controlled for age (in years), gender (1 = male, 2 = female), education level (1 = "high school", 2 = "institute of technology", 3 = "bachelor", 4 = "master's", 5 = "doctorate"), tenure (in years) and job type classifications (1 = "technical (R&D)", 2 = "marketing/sales", 3 = "administrative", 4 = "financial/accounting", 5 = "managerial", 6 = "other"). At the team level, we controlled for team size (total number of team members), team age (in years) and leader tenure within the team (in years). We also controlled for transformational leadership, as we suggested at the outset that entrepreneurial leadership is more facilitative than transformational leadership to enhance creativity. For those purpose, we used a seven-item measure on a five-point scale from Carless et al. (e.g. "My leader treats staff as individuals and supports and encourages their development").

Data Aggregation

We tested whether statistically aggregating data from employee responses to team-level constructs would be justified. Specifically, we computed the within-group interrater agreement (r_wg) and intra-class correlation (ICC) (Bliese 2000). The ICC(1) values of entrepreneurial leadership, creative team efficacy and transformational leadership were 0.29, 0.26 and 0.17, respectively, while the ICC(2) values were 0.69, 0.65 and 0.53, respectively (all ps < .001). Moreover, the mean r_wg values of entrepreneurial leadership, creative team efficacy, and transformational leadership were all above 0.95. The results indicate that aggregation is justified.

Validity Analyses

To assess the discriminant validity of the measures in our study, we conducted a confirmatory factor analysis (CFA) for entrepreneurial leadership, creative self-efficacy, team creative efficacy and employee creativity (Anderson and Gerbing 1988). The results are presented in Table 1. The proposed five-factor model demonstrated a better fit to the data (χ² [191] = 368.85, p < .001, CFI = 0.94, RMSEA = 0.06, IFI = 0.94, TLI = 0.93, RMR = 0.04) than the following alternative models. These results provide support for the distinctiveness of the four study variables for subsequent analyses.

Table 1 Results of the CFA models

CFA models	χ2/df	CFI	RMSEA	IFI	TLI	RMR
5-factor model: Baseline model	368.85/191	0.94	0.06	0.94	0.93	0.04
4-factor model:
Combine entrepreneurial leadership and team creative efficacy	990.71/246	0.77	0.11	0.77	0.74	0.07
Combine team creative efficacy and creative self-efficacy	920.05/246	0.79	0.11	0.79	0.77	0.09
Combine employee creativity and team creativity	997.34/249	0.77	0.11	0.77	0.75	0.09
3-factor model: Combine entrepreneurial leadership, team creative efficacy, creative self-efficacy	1195.10/249	0.71	0.13	0.71	0.68	0.07
1-factor model: Combine all variables	1579.63/252	0.59	0.15	0.59	0.53	0.07

Analytic Strategy

Given the proposed relationships from multilevels, we used hierarchical linear modeling (HLM) and hierarchical regression analysis to test our hypotheses. Specifically, to test hypotheses (H) 1, 3, 4, 6 and 7, we conducted three-step regressions: (1) The independent variable (entrepreneurial leadership) should be significantly related to the dependent variable (employee creativity), which tests H1. (2) The independent variable should be significantly related to the mediating variable (creative self-efficacy), which tests H3. (3) The mediating variable should be related to the dependent variable with the independent variable included in the equation, which tests H4 and H7. Moreover, to examine the relationship between team creative efficacy and employee creativity (H6), we regressed team creative efficacy at level 2 on employee creativity at level 1 in the HLM. At the team level, to test hypotheses 2, 5 and 8, we conducted hierarchical regression analysis because all the variables (entrepreneurial leadership, team creative efficacy and team creativity) are at the team level. Finally, we used the Monte Carlo method to estimate the confidence intervals (CIs) of indirect effects. The analyses were conducted after we mean-centred all the variables.