Absence of a meaningful effect of intranasal oxytocin on trusting behavior: a registered report with pooled equivalence testing

The neuropeptide oxytocin (OXT) is thought to modulate important aspects of prosocial behavior. In a seminal paper, Kosfeld et al. (2005) reported that intranasally administered OXT modulated trusting behavior in an economic trust game. Several attempts to conceptually replicate these findings yielded mixed results, which might be partly due to small sample sizes that can reduce the ability to detect, or reject, meaningful effects. We performed a large-scale replication (n = 211) of Kosfeld et al. (2005) with specific attention for small effects and subpopulations whose trusting behavior may be sensitive to OXT manipulations. Moreover, we conducted a pooled analysis of the two largest, recent, replications by merging our data with data from Declerck et al. (2020; n = 321). We found no evidence that intranasal OXT administration increases trusting behavior, neither in our data nor in the pooled dataset. While equivalence testing in our dataset was inconclusive, equivalence testing in the pooled dataset (n = 532) indicated that the effect of OXT administration on trusting behavior lies within a minimal range of effects, suggesting that the effect is too small to be of interest to (most) lab-based studies due to feasibility constraints of collecting data from larger samples required to reject smaller effect sizes. In addition, we found no evidence that OXT administration effects on trusting behavior are influenced by baseline trust, reward sensitivity, and punishment sensitivity. Our findings invite a critical evaluation of current methodology in OXT research and a reconsideration of OXT's involvement in social behaviors and beyond.