Table 10.
Endogeneity test (Hausman Wu complete results).
| Check the relationship | OLS estimation | IV estimation [c] | Difference (D) | Hausman χ2 | P | Conclusion | Diagnostic indicators |
|---|---|---|---|---|---|---|---|
| TAM → ERE | 0.57 (0.05) | 0.61 (0.06) | −0.04 | 1.27 | 0.260 | No endogeneity | KP rk LM = 12.38 (p < 0.001) |
| SSP → SFS | 0.40 (0.06) | 0.43 (0.07) | −0.03 | 0.83 | 0.362 | No endogeneity | CD Wald F = 18.92 |
| CC regulatory effect [d] | 0.24 (0.04) | 0.27 (0.05) | −0.03 | 2.15 | 0.142 | No endogeneity | Overidentification with a chi square of 0.92 (p = 0.337) |
| The first stage of regression | |||||||
| Tool variable → TAM [e] | – | 0.68 (0.09) | – | t = 7.56** | <0.001 | Strong tools | Deviation R2 = 0.32 |
| Tool variable → SSP [e] | – | 0.52 (0.07) | – | t = 7.43** | <0.001 | Strong tools | Sheas R2 = 0.28 |
Complete explanation: c instrumental variable = self-assessment of artistic literacy (correlation with TAM r = 0.51 **, correlation with error term r = 0.08, p = 0.412); the regulatory effect of d CC refers to the interaction term between TAM and CC. the first stage regression coefficient reports the impact of instrumental variables on endogenous variables.