1
$\begingroup$

I want to find out if 2 proportions are different or not.

For one, sample size is <30.

For the other, it's much >30.

Can I use the Z test since one size is large? Or do i have to use the t test? Or can I not compare since the numbers are very different? I'd be grateful for any help. Thanks.

$\endgroup$

1 Answer 1

1
$\begingroup$

As you can learn from this answer

T-values are used for inference with means because the sample standard deviation has to be used as an estimate for a population standard deviation, rather than using the population standard deviation directly (this is generally unknown).

Proportions are a little different in that the standard deviation of the hypothetical curve with mean $ p $ can be calculated by $ \sqrt{\frac{p(1-p)}{n}} $. [...]

So for proportion, the standard deviation is not unknown, which means that you should use the $Z$-test.

However, more likely you should rather use the Pearson's $\chi^2$ test (implemented in prop.test in R) or the exact binomial test (binom.test in R).

$\endgroup$

Your Answer

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

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