Work

From Bias to Brevity: Balancing Coder Bias and Team Size in Thin-Slice and Full-Length Observations of Couple Interactions

Year: 2026

Type: article

Abstract: Behavioral observation offers a powerful approach for relationship science but is constrained by coder bias and resource demands. Using a widely employed couple conflict discussion paradigm, we illustrate how these challenges can be addressed. First, we show that thin-slice coding, a method commonly used in zero-acquaintance research, can effectively substitute for full-length observations in already formed relationships. Thin-slice ratings of partner responsiveness mirrored full-length ratings, correlated with partners’ self-perceptions, and predicted changes in relationship satisfaction over time. Next, combining theory on interpersonal perception and an optimization approach in computer science, we estimated the optimal coding team size. Small teams (e.g., two coders) introduced bias and attenuated associations, but these issues were largely mitigated with five full-slice or eight thin-slice coders. Together, these findings establish thin-slice coding as a valid approach for studying close relationships and offer practical benchmarks for balancing coder bias and team size in observational relationship science.

Source: Journal of Social and Personal Relationships

DOI: 10.1177/02654075261443568

Authors: Emre Selcuk, Onur Varol, Gul Gunaydin, Zeynep Deniz Ozden, Turan Deniz Ergun

Institutions: Sabancı Üniversitesi, University of Twente

Language: English

Topic: Attachment and Relationship Dynamics

Subfield: Social Psychology

Field: Psychology

Domain: Social Sciences

Open Access status: hybrid

Estimated APC paid: $3,650

Funders: Türkiye Bilimsel ve Teknolojik Araştırma Kurumu, Sabancı Üniversitesi