Inferential Statistics - Wilcoxin Signed rank test

[frogger]

Hey guys!

I'm nearly finished my final copy for my IA- it's just that on my draft my teacher has written "not reported" for the inferential statistics part (and thus, I got a 0 for it). He's created this excel spreadsheet that gives us all the numbers we need, but I'm not sure which ones I need to put in to my report? Oh, I used the Wilcoxin Signed Rank test because I used repeated measures (serial position effect- both groups did both the control and experimental trials)

The list of err.. 'things' he's given us are:

The differences

The ranks of the differences

How many pairs we're working with (I thought this was kind of obvious? but oh well)

The sums of the positive and negative ranks

The W-value

[pick whether it's a 1 or 2 tailed test]

The 'values from the tables'

(in the description it says "There is a 5 chance of getting below 36 if there was really no difference between the categories- similarly, there's a 1% chance of getting below 24)

And our conclusion - mine unfortunately was to accept the null hypothesis but I still don't understand exactly which values I should include in my report?

I attached a screenshot of the excel spreadsheet if it helps

Thanks so much!

[blindpet]

I think you should be reporting the values like this (bear in mind the Wilcoxon test isn't used very much so my experience with reporting the values APA style is limited unfortunately). You should report the W(number of pairs)=smaller W value and the p value along with the direction of the test (one or two tailed).

Example:

A one-tailed (if you made a specific prediction about the experiment) Wilcoxon t-test was run for the two conditions (name them obviously) but the results were non-significant (W(16)=40.5, p>0.05). There was no difference between the two experimental groups (then mention later in the discussion how you had to accept the null hypothesis).

Hope this helps, your teacher should've explained how s/he wanted the results reported...