9. You wish to correlate the number of errors committed on a problem-solving task with 9.

scores on a measure of impulsivity administered to a sample of college students. Your hypothesis is that students higher in impulsivity will tend to make more errors. You wish to detect an effect of ρ 40 and have set power equal to .80.

(a) What are your statistical hypotheses?

(b) What is the required sample size to detect the specified effect at the desired level of power?

(c) Assuming a false H0, what proportion of such investigations—conducted repeatedly under identical conditions—would you nonetheless expect to result in nonsignificance?

(d) Suppose you were able to recruit only 22 volunteers for your investigation. From Table 19.2, what would be your best estimate of the power of your test?

(e) Given the situation in Problem 9d and assuming a false H0, what proportion of such investigations—conducted repeatedly under identical conditions—would you expect to result in nonsignificance?