Two-tailed or one-tailed test? The study using 139 students studies the maximum amount of alcohol consumed. Based on data are there differences between males and females in the maximum amount of alcohol consumed? (Use alpha= .05 Men Women mean=8.2 mean=5.6 s=5.9 s=5.7 n=54 n=85 I'm not sure if i should be using one-tailed or two-tailed test. I'm also not sure if my t.05 value is correct. I know you need to double the critical value column if it's two-tailed test. Am I starting this problem correct way? thank you!!enter image description here

> I'm not sure if i should be using one-tailed or two-tailed test.

Yes, a two-tailed test is correct because the maximum amount consumed by men can be more or less than that of women. Experimentally, you should establish the rejection region _before_ you collect the data, and as it is possible in theory at least that men could drink less than women, you should include that.

In general, the two-sided test is typically viewed as a better way to perform a hypothesis test, because the one-sided test makes it easier to reject the null hypothesis after you get the data by choosing the single big rejection region to be on the side the data suggests.

> I'm also not sure if my t.05 value is correct.

For $\alpha=0.05$, yes, $1.96$ is the $97.5$th percentile of the null distribution.