More on ANOVA

Answers

1. Trial, Dose, and Day. If different groups are used for each condition, then the variable is a between-subjects variable; if the same subjects are tested in each condition, then the variable is a within-subjects variable.
   
   
2. For between-subjects variables, differences among subjects are error. For within-subjects variables, each subject's performance in one condition is compared to his or her performance in another condition. Therefore, overall differences are not a source of error.
   
   
3. The assumption has to do with correlations among scores, which is not applicable in between-subjects designs since subjects have only one score.
   
   
4. Violating the assumption increases the Type I error rate, sometimes substantially.
   
   
5. 0.049