Please Help: Can I remove a level of a factor?

GuPug

New Member
#1
I need to compare the performance of two groups of subjects undergoing two different treatments over time.

I have used a two-way repeated measures mixed ANOVA (SPSS) with
- Group (treatment A vs B) as between factor
- Time (pre treatment, 1 month after treatment, 2 month after treatment) as within-factor.

The interaction between the two variables is not significant (0.35) . In fact both group A and B perform better at 1 month but at 2 months they have dramatically different outcomes (see picture 1 ): Group A's (blue) performance is better than Group B (green).

If I remove a level of one of the factors (time) using the same analysis (see picture 2): Group (2 levels) x Time (2levels), an interaction occurs (0.02).
Am I allowed to do so? From a statistical perspective, can I sustain this choice with my tutor (he already knows that the first analysis does not produce a significant result).

Thank you very much for your assistance!
 

Attachments

Last edited: