5
$\begingroup$

My independent variable has a huge proportion of zeros in the data, and I've been thinking of using a zero-inflated Poisson regression to account for this. However, this method is only specified for if the dependent variable has many zeros. Does it work if only the independent variable has the same issue? Are there any other methods to remedy this if not (like some sort of transformation on the independent variable)? Thanks.

$\endgroup$
0

1 Answer 1

5
$\begingroup$

A zero-inflated model is appropriate when the dependent variable has an excess of zeros that cannot be explained by standard models for count data such as Poisson.

There is no requirement, condition, or assumption regarding the distribution of the independent variable(s) in a regression model. We might like the distribution of the dependent variable, conditional on the independent variable(s) to be approximately normal, and for this we would normally inspect the residuals of the model.

$\endgroup$
0

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.