How to explore the baseline level effect on response size (i.e. CFB)? Is fit model the right tool in JMP?

[#1]

Hi All, 

We have a test that compared the benefit (data reading, numerical data) of two different treatments over a couple of time points (baseline, T1, T2, etc). There are also some other effects: test sites, gender, etc.

For primary objective, I used the "Fit model" and put the data reading as Y, treatment, baseline level, time point, gender, subject ID(random factor), some 2-factor interactions in the model and hit run. Then see if treatment is a sig effect overall and at each time point.

We notice the response size (i.e. change from baseline CFB) was smaller in data from test site A than test site B. And test site A baseline level mean is higher than that of test site B. So I want to understand from the data if sample with lower baseline would response more/ have a bigger CFB?

I repeat the Fit model, and put CFB in Y. The effect summary showed "baseline level" effect is sig as well as baseline level*treatment interaction.  But when looking into the effect details, there's nothing to show under baseline level effects. All the items plot, student T are greyed. I couldn't learn if higher baseline leads to smaller or bigger CFB. 

How to best explore this question and have some more detailed results/plots than just a p?

Thank you all

Diana

[#2]

Hi Diana,
Would it be possible to see a snapshot of your model output (you can hide any confidential / sensitive information prior to sharing).
At first glance, the leverage plot should give you the directionality and magnitude of the effect. I also find the Profiler to be a useful tool to understand the relationships between outcome and variables.
Best,
TS

[#3]

Hi Thierry, Thank you for reach out.

There are the snapshot of the report parts that relevant to baseline level. All items are not available in the profilers of baseline level related effect. I don't have a clue of why.