• Welcome! The new FreeMathHelp.com forum is live. We've moved from VB4 to Xenforo 2.1 as our underlying software. Hopefully you find the upgrade to be a positive change. Please feel free to reach out as issues arise -- things will be a little different, and minor issues will no doubt crop up.

Drugs testing, which t test


Junior Member
Jan 26, 2012
I am doing an investigation into the effects of drug X on 20 patients. I have the results of their test scores for before and after their dosage.

My H[SUB]0[/SUB]: There is an effect of the drug
vs H[SUB]1[/SUB]: No effect

What t-test do I use? I think my test statistic is

t= [x[SUB]1(bar) - [/SUB]x[SUB]2(bar)[/SUB]] / s.e.[x[SUB]1(bar) - [/SUB]x[SUB]2(bar)[/SUB]] where the s.e. is sqrt[(s[SUB]1[/SUB][SUP]2[/SUP]/n[SUB]1) +[/SUB](s[SUB]2[/SUB][SUP]2[/SUP]/n[SUB]2[/SUB])] where n is the number of patients in each sample (in this case 20 for each), and s[SUP]2[/SUP] are the sample variances

and my t value is t~t[SUB](n1+n2-2)[/SUB], just to clarify, using this test makes the assumption that the population variances are the same.

Is this the correct t-test to use?
Last edited: