How to tell (know) if an explanatory variable has impact or not and how much?
To test the explanatory power of the whole set of explanatory variables, as compared to just using the overall mean of the outcome variable, use the F-statistic and the p-value printed by SPSS or Excel under “ANOVA.” If this p-value is less than 0.05, you can reject the null hypothesis (which is that all of the variables have coefficients of zero) at the 5% significance level. If the p-value is less than 0.01 you can reject at the 1% significance level.
Comments
Leave a comment