Aggregating to the between-person level in idiographic research designs: Personal goal research as an example of the need to distinguish between reliability and homogeneity

Abstract

Abstract In several areas of personality research (e.g., personal networks, interpersonal behaviour, academic effort), the empirical data represent a multilevel structure, in which within-person variables are nested within individuals. Both within-person variation and between-person differences can be related to outcome variables (e.g., emotions, life satisfaction). One major precondition for aggregating within-person data to the between-person level is sufficient reliability of the aggregated data, a point that has received scant attention in previous research. Drawing on the example of personal goal research, we recommend that the two major forms of the intraclass correlation coefficient—the ICC(1) and the ICC(2)—be applied to within-person data in idiographic research designs. These indices allow the homogeneity and reliability of within-person data to be determined separately. Their application is illustrated using data from a study with 4,565 young adults.