[ricogallo]

Yes, traditionally you need the 5% of error, or even 1%. Also the partial eta-squared is the ANOVA measure for the size of the effects studied, which happens to be very low too: 9%.