According to one of my research hypotheses, personality characteristics are supposed to influence job satisfaction, which are gender+Age+education+parenthood, but when checking for normality and homogeneity of the dependent variable(job sat,), it is non-normally distributed for gender and age. Am i supposed to exclude age and gender from the model, should i find non-parametric alternative, or should i conduct linear regression anyway?
What Sofia is saying is that the residuals must be normal not Y. Best Wishes
I agree with sofia that The assumptions for linear regression are that the error terms are independent and normally distributed with equal variance.
so first check this assumption. If it doesn't fulfill then go for Non-parametric alternative.
one more thing As your dependent variable seems to be categorical, you can go for Multinomial logistic regression also.
- Badges
- Science topic
- Similar topics
- Engineering
- Control Theory
- Control Systems
- Control Systems Engineering