# Can I average samples in nested factor within-subject design if samples are from the same condition?

This question relates to one of my other questions: What can cause a "Error() model is singular error" in aov when fitting a repeated measures ANOVA? .

I have subjects taking an experiment, where I have different Scenarios and each Scenario contains some number of Trials, where subjects can interact with each other between Trials, but the Scenario stays the same. Can I consider each Scenario as a different treatment condition and the trials as samples so I can compute the average on the Trials of each subject for each Scenario?