In Poisson and negative binomial regression, the response is assumed have Poisson and negative binomial distributions respectively. When we test the significance of the parameter $\beta$, which is estimated by maximum likelihood estimation, we have the Wald chi-squared statistic

$\left(\frac{\beta}{\text{standard error}(\beta)}\right)^2$.

The questions are:

  1. How to derive this Wald chi-squared statistic from the response distribution assumption?
  2. How to derive the standard error for this case?

1 Answer 1


The results are more general than the ones asked of this question. The tests are generally used for most Generalized Linear Models (your specification choses two particular link functions i.e., functions describing the connection between the mean of a distribution to a predictor variable). The proofs for the most general case can be found in any standard text on large sample theory (see for instance Pg 508-510 in 'Testing Statistical Hypothesis' by Lehmann and Romano).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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