How to interpret this data.

[#1]

I have a question to the JMP-community regarding how I should interpret these data. I use JMP Pro 16.0.0 and is very satisfied with it. However, my statistician uses SPSS and now a question has arrived as to how I should interpret these data which are enclosed in this PDF-file.

Picture 1: I have participants randomised into two groups: Experiment and Control. Each of the participants perform a pre-test and then a post-test. The values are in percent. When I want to study if the post-test values have increased as compared to the pre-test I have used the paired t-test for matched variables, since then each individual is his/her own control.

Picture 2: What I interpret is that in both groups the post-test values have increased significantly.

Picture 3: What I then wanted to study is whether there is a significant difference in the rate of improvement between the two groups. Of course my wish is that the experiment group has improved more than the control group. In order to study this I put the Group differentiator in the X, grouping box.

Then is the question how should I interpret the results. My interpretation from the results in picture 3 is that the experiment group has improved significantly more than the control group. Is that a correct interpretation?

[#2 · Accepted solution]

 Your second picture (with Group in the By Box) simply does the matched pairs test for each group separately.  The numbers certainly suggest that the experimental group achieved greater gains than the control group, but that analysis window does not provide a test of that.  Your third picture (with Group in the Group By box) appears to provide a sort of weighted average of the 2 groups - both of which show improvements from pre to post test.

I'd suggest simply creating the variable showing the difference in scores from pre to post test and doing Fit Y by X, with Y as the difference and X as the Group.  That should directly address your question.  Matched pairs are really a different way of looking at a single variable analysis where the variable is the difference in scores.  It is more straightforward to just look at the score difference in relation to the groups.