Quantitative Methods Research

Quantitative Methods psychologists at USC conduct research on five interrelated topics.

Robust Statistical Inferences. Research on robust inferences strives to understand the most reliable statistical features of any behavior. This work includes a complete revision of univariate and multivariate statistics from the direct identification of outliers and influential observations. This topical area is led by Rand Wilcox.

Psychometrics and Measurement. The basic principles of psychological measurement are used and applied to scale development in several content areas. This includes the uses of recent advances in Item Response Theory (IRT) and in Common Factor Analysis (CFA) are applied to scale development, and form the basis of computer adaptive testing (CATI). This topical area is led by John McArdle.

Behavior and Molecular Genetics. Individual differences in psychological behaviors are a complex function of both genetic and non-genetic sources, and advances in statistical analysis have played a crucial role in new results. We examine the basic benefits and limitations of family data, including twins, and these issues are combined with the use of measured genotypes to better understand these sources of variation. This topical area is led by Laura Baker and Carol Prescott.

Decision Making in Real Life. Individuals and groups make many important real-life decisions about health and work, marriage and family planning, and about engaging in risky behaviors. We study the elementary processes behind such decisions, including the development of group and individual utility functions. This topical area is led by Richard John.

Longitudinal Dynamic Changes. The accurate measurement of developmental changes from longitudinal data are a mainstay of developmental, personality, and motivational psychology. Recent advances in latent trajectory analysis, multi-level survival analysis, growth mixture modeling, and systems dynamics modeling, are all combined with contemporary psychometric measurement models. This topical area is led by John McArdle.