Create a scatterplot of the data in Table 6.6 and note that six points having X values greater than 125 are visibly separated from the bulk of the observations. Now compute a 0.95 confidence interval for the slope of the least squares regression line with the six points having X values greater than 125 eliminated. The resulting confidence interval is

(−0.84, −0.17), so you reject the hypothesis of a zero slope and conclude that the two variables are dependent. Comment on this result in the context of factors that affect the magnitude of Pearson’s correlation which were described in Section 6.4.