All articles report statistical methods, such as ordinary least square regression, the inclusion of firm and industry fixed effects in the regression analysis, path analysis, duration model, and structural equation modeling. Nine articles adopt procedures to correct the consequences of heteroskedasticity for ordinary least squares estimation. Fourteen articles assess the potential multicollinearity problem by using correlation coefficients or the variance inflation factor. Fifteen articles provide sensitivity or additional analysis to examine the robustness of their research results. As to the omitted variables, most studies include in the analysis control variables suggested in prior literature. Overall, the research hypotheses are developed from the relevant theories and prior literature, and the statistical analyses are performed appropriately; both are helpful in ensuring internal validity.
All articles report statistical methods, such as ordinary least square regression, the inclusion of firm and industry fixed effects in the regression analysis, path analysis, duration model, and structural equation modeling. Nine articles adopt procedures to correct the consequences of heteroskedasticity for ordinary least squares estimation. Fourteen articles assess the potential multicollinearity problem by using correlation coefficients or the variance inflation factor. Fifteen articles provide sensitivity or additional analysis to examine the robustness of their research results. As to the omitted variables, most studies include in the analysis control variables suggested in prior literature. Overall, the research hypotheses are developed from the relevant theories and prior literature, and the statistical analyses are performed appropriately; both are helpful in ensuring internal validity.
正在翻譯中..
