Between-Reporter Agreement Among Couples: Exploring Nonlinear Patterns in Marital Conflict

This study compares self-reported and spouse-reported frequencies of various types of behaviors couples use in marital conflict using data from Turkish Family Structure Survey (2016), which included 9,636 couples. Existing literature reported low levels of agreement among couples. These studies measured between-reporter agreement levels among couples with linear models which report cumulative scores. However, we hypothesized that low levels of agreement reported in the literature might be due to nonlinear patterns in the self-reported and spouse-reported frequencies. To detect such nonlinear patterns, we compared self-reported and spouse-reported frequencies for each level (frequency) of using the types of conflict behavior styles we investigated. This allowed us to explore the patterns of agreement or disagreement at the low, medium, or high levels of using these conflict style behaviors. Findings of this study confirmed our hypothesizes. There are meaningful nonlinear patterns that are not detectable with cumulative comparative scores. There are also differences in these patterns across the types of conflict behaviors we looked at. For example, the nonlinear patterns we observed for aggressive conflict behavior types are different from the patterns for passive conflict behavior types. We also found that these patterns are different for husbands and wives. This paper highlights the limited capacity of linear models for exploring between-reporter agreement levels and it calls for the consideration of using nonlinear methods for these types of systematic investigation. The major limitation is the lack of analysis on the effect of sociodemographic factors on the observed discrepancies other than gender.