Say for example, there is an observed estimated standardized mean difference of $d = 0.44$ with $CI_{95} = [-.10, .89]$. This is the only information we have available (i.e. we do not know the sample size of either comparison group, or any sample statistics). With this information alone, is it possible to determine if the results are significant given $\alpha = .05$ (two-tailed)?
1 Answer
$\begingroup$
$\endgroup$
Presumably your null hypothesis was $H_0:\mu_1-\mu_2=0$. So you have your difference of means there and your CI crosses the value of 0, which is the same as the proposed null value, hence your observed difference is not inconsistent with the null. Do not reject.
