Results of robustness tests
| Model | (1) Public | (2) Public | (3) Public | (4) Public | (5) Public |
|---|---|---|---|---|---|
| Disposition | −0.0777 | −0.127* | −0.164* | −0.165* | −0.0817*** |
| (0.0615) | (0.0624) | (0.0796) | (0.0787) | (0.00945) | |
| Work | −0.499*** | −0.521*** | −0.757** | −0.724** | −0.505*** |
| (0.0238) | (0.0240) | (0.267) | (0.269) | (0.0226) | |
| Richness | 0.936*** | 0.922*** | 0.969*** | 0.970*** | 0.886*** |
| (0.0162) | (0.0163) | (0.0169) | (0.0169) | (0.0158) | |
| Disposition* Work | 0.0667*** | 0.232** | 0.214* | 0.0605*** | |
| (0.0122) | (0.0880) | (0.0877) | (0.0112) | ||
| Disposition* Richness | 0.0646*** | 0.0635*** | 0.0636*** | 0.0607*** | |
| (0.00833) | (0.00861) | (0.00861) | (0.00793) | ||
| Benefit | 3.103*** | 3.092*** | 3.162*** | 3.176*** | 3.468*** |
| (0.0862) | (0.0862) | (0.0894) | (0.0895) | (0.0792) | |
| Risk | −0.0347 | −0.0570 | 0.239 | −0.0119 | −0.256*** |
| (0.192) | (0.194) | (0.248) | (0.222) | (0.0195) | |
| Gender | −0.370 | −0.361 | −0.615 | −0.558 | −0.139*** |
| (0.286) | (0.288) | (0.369) | (0.320) | (0.0318) | |
| Age | −0.456*** | −0.459*** | −0.555*** | −0.545*** | −0.245*** |
| (0.0341) | (0.0341) | (0.0357) | (0.0357) | (0.0114) | |
| Toi | 0.0000868 | 0.0000896 | 0.0000576 | 0.0000905 | 0.0000436*** |
| (0.0000624) | (0.0000629) | (0.0000806) | (0.0000697) | (0.00000244) | |
| Experience | 2.721*** | 2.706*** | 2.262*** | 2.647*** | 2.341*** |
| (0.274) | (0.276) | (0.355) | (0.315) | (0.0251) | |
| Constant | −2.958*** | −2.946*** | −2.799*** | −2.920*** | −2.659*** |
| (0.262) | (0.264) | (0.338) | (0.317) | (0.0248) | |
| Log-likelihood | −32686.842 | −32655.151 | −31235.26 | −31226.457 | −34145.178 |
| Likelihood-ratio test | 63.38*** | 2839.78*** | 17.60*** | ||
| Pseudo-R2 | 0.141 | 0.142 | 0.179 | 0.179 | 0.254 |
| AIC | 65395.68 | 65336.3 | 62498.52 | 62482.91 | |
| BIC | 65499.33 | 65458.79 | 62630.43 | 62624.25 |
| Disposition | −0.0777 | −0.127* | −0.164* | −0.165* | −0.0817*** |
| (0.0615) | (0.0624) | (0.0796) | (0.0787) | (0.00945) | |
| Work | −0.499*** | −0.521*** | −0.757** | −0.724** | −0.505*** |
| (0.0238) | (0.0240) | (0.267) | (0.269) | (0.0226) | |
| Richness | 0.936*** | 0.922*** | 0.969*** | 0.970*** | 0.886*** |
| (0.0162) | (0.0163) | (0.0169) | (0.0169) | (0.0158) | |
| Disposition* Work | 0.0667*** | 0.232** | 0.214* | 0.0605*** | |
| (0.0122) | (0.0880) | (0.0877) | (0.0112) | ||
| Disposition* Richness | 0.0646*** | 0.0635*** | 0.0636*** | 0.0607*** | |
| (0.00833) | (0.00861) | (0.00861) | (0.00793) | ||
| Benefit | 3.103*** | 3.092*** | 3.162*** | 3.176*** | 3.468*** |
| (0.0862) | (0.0862) | (0.0894) | (0.0895) | (0.0792) | |
| Risk | −0.0347 | −0.0570 | 0.239 | −0.0119 | −0.256*** |
| (0.192) | (0.194) | (0.248) | (0.222) | (0.0195) | |
| Gender | −0.370 | −0.361 | −0.615 | −0.558 | −0.139*** |
| (0.286) | (0.288) | (0.369) | (0.320) | (0.0318) | |
| Age | −0.456*** | −0.459*** | −0.555*** | −0.545*** | −0.245*** |
| (0.0341) | (0.0341) | (0.0357) | (0.0357) | (0.0114) | |
| Toi | 0.0000868 | 0.0000896 | 0.0000576 | 0.0000905 | 0.0000436*** |
| (0.0000624) | (0.0000629) | (0.0000806) | (0.0000697) | (0.00000244) | |
| Experience | 2.721*** | 2.706*** | 2.262*** | 2.647*** | 2.341*** |
| (0.274) | (0.276) | (0.355) | (0.315) | (0.0251) | |
| Constant | −2.958*** | −2.946*** | −2.799*** | −2.920*** | −2.659*** |
| (0.262) | (0.264) | (0.338) | (0.317) | (0.0248) | |
| Log-likelihood | −32686.842 | −32655.151 | −31235.26 | −31226.457 | −34145.178 |
| Likelihood-ratio test | 63.38*** | 2839.78*** | 17.60*** | ||
| Pseudo- | 0.141 | 0.142 | 0.179 | 0.179 | 0.254 |
| AIC | 65395.68 | 65336.3 | 62498.52 | 62482.91 | |
| BIC | 65499.33 | 65458.79 | 62630.43 | 62624.25 |
Note(s): *p < 0.05, **p < 0.01, ***p < 0.001. [1] Standard errors are in parentheses. [2] The regression method used for models (1) to (4) is meqrlogit and model (5) uses GSEM. [3] The first part of Table 3 reports the coefficients (standard errors) and the significance levels of the fixed effects and the second part reports the results of model comparisons. [4] The likelihood-ratio test values for models (2), (3) and (4), respectively, correspond to (1), (2) and (3), indicating that the model is improving incrementally. [5] The Pseudo-R2 of the model (5) is given based on the constant term model of GSEM: (45769.576–34145.178)/45769.576 = 0.254
Source(s): Authors' own creation/work
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.