1 breifly explain the statistical advantage withinsubject de

1. breifly explain the statistical advantage within-subject desgins have over between subject designs.

2. the SAT is a very good predictor of first semester GPA in college. when conducting a study with Princeton freshmen, however they found that SAT had no relationship at all with GPA. WHY?

3. if i say that variable Z(attitudes towards poor people) mediates the relationship between X (education) and Y ( charitable giving to the poor) what exactly do I mean?

Solution

1. The main statistical advantage within-subject designs have over between subject designs is that fewer treatments are needed in a within-subject design since each treatment is tested on all levels of a factor. Although more testing is required for each treatment, but overall testing is easier for fewer treatments. Also, since there are less treatments in within-subjects designs, there is less variance due to treatment disposition. In within-subject designs, since treatments take part in all conditions, they allow us to control for many inter-individual confounding variables.

2. SAT is a good predictor of first semester GPA in college as both is a measure of I.Q. However, SAT has no relationship at all with GPA as both as measured on different scale, and are calculated with different criterias and guidelines.

3. By saying that Z (attitude towards poor people) mediates the relationship between X (education) any Y (charitable giving to poor), we mean that Z explains how X is actually related to Y. If we say education is related to charity, but we do not know how. But, if we say that education makes the attitude and attitude, in turn, is related to charity, hence explaining the relationship between education and charity.