when the size of the difference between group means in an experiment is small compared to the amount of variability within each condition, whcih of the following is likely true? 1) effect size d is large 2) effect size d is small 3) the effect is significant 4) there is a design confound