3
$\begingroup$

I have a list of insurance policies that filed claims in pre/post periods. I am trying to test whether the average cost of insurance claims per policy varied. Some of my policies don't have any claims in pre, but have some claims in post.

Should I make the average claim cost of these policies zero or should I ignore the whole policy? Or something else?

$\endgroup$
1
  • 1
    $\begingroup$ Do you have any other variables/covariates available to you that might enable to you to predict the missing values? $\endgroup$ Apr 24, 2015 at 19:44

1 Answer 1

1
$\begingroup$

If you want to check whether the average of pre and post claims cost of insurance claims per policy varied for claim(s) rise, then you should ignore.

&

If you want to check whether average of pre and post claims cost of insurance claims per policy varied for policies exposed to risk,then you should not ignore.

Then, use Paired t-test.

$\endgroup$

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.