Where the Choice of Model Leads Us: An Empirical Comparison of Dyadic Data Analysis Frameworks

Abstract

AbstractDyadic data require unique approaches due to their innate ‘interdependence’. The Actor-Partner Interdependence Model (APIM; Kenny, J Soc Pers Relat 13: 279–294, 1996) and Common Fate Model (CFM; Kenny and La Voie, J Pers Soc Psychol 48(2): 339–348, 1985) and their extensions are some of the most widely used dyadic data analysis approaches. They have been applied to study various dynamics among dyads, such as couples, twins, and matched pairs (e.g., Maroufizadeh et al., BMC Med Res Methodol 18(1): 1–10, 2018; Moorman, 2016). Whereas the APIM framework specifies explicit and direct paths from members to themselves and their partners, the family of CFM views interdependence as arising from a joint process within each dyad rather than as a sum of individual influences. Despite these conceptual differences, both approaches have been utilized in similar fields of research, and their unique advantages and limitations remain to be understood.This study examines the two families of dyadic data analysis models, APIM and CFM, with respect to their conceptual differences as well as their empirical implications. An illustrative example using the Wisconsin Longitudinal Study data demonstrates how research questions addressed with different specifications of dyadic data analysis models lead to distinguished conclusions. Following discussions include future research directions for studying dynamic relationships and interdependence in dyadic and other small group data in the fields of psychology and education.KeywordsDyadic data analysisActor-partner interdependence modelCommon fate modelWisconsin Longitudinal Study data