Comparison of Dependent Spearman Correlation Coefficients

• Activity Number: 124
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Bernard Rosner*+ and Wei Wang and Heather Eliassen and Eileen Hibert
• Companies: Harvard Medical School and Channing Division of Network Medicine and Channing Division of Network Medicine and Channing Division of Network Medicine
• Keywords: Spearman correlation ; dependent correlations ; deattenuated correlation ; measurement error ; probit score

Abstract

A marker of validity of a dietary instrument is how well it correlates with a relevant biomarker. For example, there are several dietary instruments for assessing dietary carotenoids, and a natural question is which measure correlates most strongly with the associated plasma carotenoid. In addition, the distributions of dietary carotenoids are often skewed and the Spearman correlation coefficient is a more appropriate measure of association than the Pearson correlation. Thus, we propose methods for comparing dependent Spearman correlations estimated from the same subjects. Also, since dietary carotenoids are measured with error, we propose on extension of these methods to compare dependent deattenuated Spearman correlations. The methods are illustrated with (a) the comparison of the validity of different methods of measuring dietary carotenoids and (b) the comparison of the validity of different storage protocols in processing plasma samples for the determination of HgbAlc.