topic Re: wilcoxon two sample test in SAS Studio

wilcoxon two sample test

laurenhosking — Mon, 11 Nov 2019 20:35:00 GMT

How would I analyse this table to know wether to reject the null hypothesis at sig level 0.05. All examples online have an extra column

Re: wilcoxon two sample test

Reeza — Mon, 11 Nov 2019 20:51:58 GMT

It's a two sample test, you need to have two columns one for your grouping variable and one for you values, so Class is your grouping variable and Grade is your continuous.
What is the hypothesis you're testing?

Re: wilcoxon two sample test

laurenhosking — Mon, 11 Nov 2019 21:01:46 GMT

the hypothesis im testing is that that the morning class performed worse than the afternoon class

morning class is 1 and afternoon is 2

Re: wilcoxon two sample test

Reeza — Mon, 11 Nov 2019 22:40:33 GMT

Your N is too small for any conclusive results, assuming this is homework, you need to run the ANOVA and interpret the results. You should have all you need to answer your question now.

Re: wilcoxon two sample test

PGStats — Mon, 11 Nov 2019 23:28:01 GMT

If your hypothesis only goes one way, i.e. it was excluded from the start that the morning class could do better than afternoon class, consider the unilateral test p-value under "Pr < Z". But if both outcomes were possible, look at the Pr > |Z| prob.

Those results are approximate (but usually pretty good). You may also request an EXACT test, if available in your version of SAS.

Re: wilcoxon two sample test

laurenhosking — Tue, 12 Nov 2019 00:17:39 GMT

Is this the Pr<z under the t approximation column? Also as I did this for a 2 sided test would I divide this answer by 2 for a one sided test. Thank you!