Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

Share this Page URL
Help

9 Correlation and Tests of Independence > 9.7 Exercises - Pg. 468

468 Introduction to Robust Estimation and Hypothesis Testing Repeat this process 599 times yielding D 1 , . . . , D 599 . Then a 0.95 confidence interval for the difference between the population correlation coefficients ( 1 - 2 ) is (D ( ) , D (u) ), where for = 7 and u = 593 if N < 40; = 8 and u = 592 if 40 N < 80; = 11 and u = 588 if 80 N < 180; = 14 and u = 585 if 180 N < 250; = 15 and u = 584 if N 250. 9.6.2 Comparing Robust Correlations When comparing robust correlations, all indications are that a basic percentile bootstrap method performs reasonably well. That is, no modification, as described in the previous section, is necessary. 9.6.3 R Functions twopcor and twocor The R function