I have to analyze two samples (n1=10,n2=18) in a design in which there is a between-subject factor (Groups: 2 levels) and a within-subjects factor (Stimuli: 2 levels). I'm wondering if, in such situations, a mixed (between/within) ANOVA is justified, given the small and unbalanced samples size. The rule of thumb for ANOVA is at least 30 subjects for cell, I know, but often I read papers in which ANOVA is used for smaller samples.

Could anyone suggests me the best practice for this situation? Can such analysis in this context be defended in a publication process?