Using the data on self-enhancement given in Exercise 10.16, add two scores to each year, keeping the year mean unchanged. (For example, if the 1982 mean SelfEn score was 0.40, you could make two other scores for that year by adding and subtracting .03 from 0.40.)

This will give you pairs of scores. Now compute the correlation and regression of SelfEn predicted from Year. Why does the correlation differ from the computed in Exercise 10.16? What has happened to the slope, and why?