a school counselor wants to compare the effectiveness of an online sat preparation program with an in-person sat preparation class. for an experiment, the counselor recruits 30 students who have already taken the sat once. the response variable will be the improvement in sat score. select all true statements. the completely randomized design is preferred the completely randomized design is better because it eliminates bias by considering all students to be equal at the start of the experiment. the matched pair design is preferred students who did not score well the first time they took the exam may benefit greatly from the class. by matching students by initial score, we can account for the variability in improvement that is introduced by the variability in student ability, making it easier to determine which program is more effective. students who scored well on the exam initially may not have much room for improvement even if the treatments are effective.