2
$\begingroup$

How can you refer to insignificant binary dummy variables in the results chapter of papers?

E.g. when it has been found that gender is not significant. Can you state that there are no significant differences between males and females regarding the target variable? Or is this not correct from a statictical point of view?

$\endgroup$
4
$\begingroup$

You should say, that you could find no evidence (or better yet, "no strong evidence") for statistically significant differences in the average outcomes between males and females at the $X$% confidence level, when controlling for the other variables in the model. Certainly there are significant differences: their genitalia for example and other minor variations that don't matter to you. It also does not make sense to talk about statistically significant differences if you don't refer to the Probability of making a Type I error somehow.

$\endgroup$
2
$\begingroup$

Your statement is nearly correct. You would interpret the non-significant effect of a dichotomous Gender variable by saying the dependent variable does not differ between the two levels of Gender when adjusting for the other variables in the model.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.