Why does ez warn: Collapsing Cell Means

[Blaine Lewis]

Hi, 

I have data that looks like: 

participant, condition, block, completed_rate

1,cond_1,1,0.05

1,cond_1,2,0.75

...

8,cond_2,3,0.6

I'm trying to find out if condition has an effect on completed_rate. Blocks represent repetitons, I analysed block and found that for the last 3 blocks it doesn't have an effect on completed_rate. 

I wanted to analyse all of the last blocks at once to get more data points and more power so after filtering to just the last four blocks I used:

m <- ezANOVA(data = df, 

      wid = participant, 

      dv = completed_rate, 

      between = condition)

However ezANOVA complains "Collapsing data to cell means." and I have no main effect. If I use aov for the same analysis and data then I get a significant result. Am I doing something wrong?