0
$\begingroup$

Let us assume that we havo two parameters $\alpha$ and $\beta$ and the two maximum likelihood estimators $\hat{\alpha}$ and $\hat{\beta}$. We have also the confidence interval of these estimators, standard error and the covariance is known. How can I derive the confidence interval of the difference $\hat{\alpha}$ - $\hat{\beta}$? I think there is a method called Delta method for this, but I'm not sure and I don't know well how it works. Thank you for anyone able to help me

$\endgroup$
3
$\begingroup$

Given MLEs $\hat{\alpha}$ and $\hat{\beta}$ and their corresponding standard errors, $\mbox{SE}[\hat{\alpha}]$ and $\mbox{SE}[\hat{\beta}]$, approximate 95% Wald-type CIs for $\alpha$ and $\beta$ can be obtained with $$\hat{\alpha} \pm 1.96 \mbox{SE}[\hat{\alpha}]$$ and $$\hat{\beta} \pm 1.96 \mbox{SE}[\hat{\beta}].$$ If you also know the covariance between $\hat{\alpha}$ and $\hat{\beta}$ (let's denote this $\mbox{Cov}[\hat{\alpha}, \hat{\beta}]$), then an approximate 95% Wald-type CI for $\alpha - \beta$ can be obtained with $$(\hat{\alpha} - \hat{\beta}) \pm 1.96 \sqrt{\mbox{SE}[\hat{\alpha}]^2 + \mbox{SE}[\hat{\beta}]^2 - 2\mbox{Cov}[\hat{\alpha}, \hat{\beta}]}.$$

$\endgroup$

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.