*4.2. Confirmatory Factor Analysis*

A confirmatory factor analysis (CFA) was adopted to ensure that the four main constructs had good discriminant validity using MPLUS 7.0. Following the practice of Keem et al. [69], a four-factor model, two three-factor models, a two-factor model, and a single-factor model were included in the confirmatory factor analysis. Results showed that the four-factor model (environmental corporate social responsibility, shared vision capability, resource slack, and green innovation performance) fit the data well: χ2/df = 3.421, CFI = 0.925, TLI = 0.912, SRMR = 0.047, and RMSEA = 0.083. For one three-factor model, we loaded environmental corporate social responsibility and green innovation performance indicators on a factor, and the results were χ2/df = 7.936, CFI = 0.780, TLI = 0.748, SRMR = 0.115, and RMSEA = 0.141. For the other, the study loaded shared vision capability and resource slack indicators on a factor, and the results were χ2/df = 6.730, CFI = 0.818, TLI = 0.792, SRMR = 0.099, and RMSEA = 0.128. In the two-factor model, we loaded shared vision capability, green innovation performance, and resource slack indicators on

a factor, and the results were χ2/df = 13.099, CFI = 0.611, TLI = 0.560, SRMR = 0.155, and RMSEA = 0.186. As stated above, for a model in which all four constructs were set to load on a single factor, the results were χ2/df = 17.073, CFI = 0.480, TLI = 0.415, SRMR = 0.171, RMSEA = 0.214. As the results show, the research model was acceptable and significantly better than the measure models.
