Which statistical test should i use?
I want to determine if your PSAT score is a good predictor of your SAT score. I sampled 100 student's PSAT and SAT scores. Now i dont know if i should use a paired t test, a 2 sample t test, or a mean difference test.
You need to think about randomization before you draw the sample. You could take the students in the 3 AP classes as the population, then use Excel to sort them according to a random number generator "=RAND()" and then use the 100 students that sort to the top of the list.