DeFries-Fulker (DF) analysis is a regression model used in behavior genetics (DeFries & Fulker, 1985) to estimate ACE (and other) models. It is appealing for its simplicity, however it violates certain regression principles including homogeneity of variances and independence of errors. These violations make calculation of standard errors and confidence intervals problematic. Methods have been developed to account for this (Kohler & Rodgers, 2001), although the research on these methods is sparse. The univariate bootstrap is a relatively recently developed version of the bootstrap (Lee & Rodgers, 1998), one that resamples from the marginal univariate distributions rather than the bivariate/multivariate data space. Currently, research on the univariate bootstrap has largely focused on individual bivariate correlations, however the univariate bootstrap represents a unique means of obtaining confidence intervals for DF models (one that is presaged by suggestions from previous DF research; e.g., Cherny, Cardon, Fulker, & DeFries, 1992). This project presents the results of a simulation study examining various methods for obtaining confidence intervals for DF model parameters, including standard confidence intervals, robust Huber-White confidence intervals, and confidence intervals from traditional and univariate bootstrapping. Results for both normal and highly skewed data are presented. It is predicted that the univariate bootstrap may potentially improve on both parametric methods and other bootstrap procedures.
Cherny, S. S., Cardon, L. R., Fulker, D. W., & DeFries, J. C. (1992). Differential heritability across levels of cognitive ability. Behavior Genetics, 22(2), 153–162.
DeFries, J. C., & Fulker, D. W. (1985). Multiple regression analysis of twin data. Behavior Genetics, 15(5), 467–473.
Lee, W.-C., & Rodgers, J. L. (1998). Bootstrapping correlation coefficients using univariate and bivariate sampling. Psychological Methods, 3(1), 91.
Kohler, H.-P., & Rodgers, J. L. (2001). DF-analyses of heritability with double-entry twin data: Asymptotic standard errors and efficient estimation. Behavior Genetics, 31(2), 179–191.