Loading evaluation data…
The neurohormone Oxytocin (OT) has been one of the most studied peptides in behavioral sciences over the past two decades. Many studies have suggested that OT could increase trusting behaviors. A previous study, based on the "Envelope Task" paradigm, where trust is assessed by the degree of openness of an envelope containing participant's confidential information, showed that OT increases trusting behavior and reported one of the most powerful effects of OT on a behavioral variable. In this paper we present two failed replications of this effect, despite sufficient power to replicate the original large effect. The non-significant results of these two failed replications clearly exclude a large effect of OT on trust in this paradigm but are compatible with either a null effect of OT on trust, or a small effect, undetectable with small sample size (N = 95 and 61 in Study 1 and 2, respectively). Taken together, our results question the purported size of OT's effect on trust and emphasize the need for replications.
two failed replications of this effect, despite sufficient power to replicate the original large effect
two randomized null studies had ample power for the originally reported effect size
The non-significant results of these two failed replications clearly exclude a large effect of OT on trust in this paradigm
Study 1 and the pooled evidence reject a large effect, though Study 2 alone was smaller
our results question the purported size of OT’s effect on trust
the evidence directly tests one paradigm, while broader interpretation relies on contextual literature
compatible with either a null effect of OT on trust, or a small effect, undetectable with small sample size
confidence intervals and power analysis preserve both null and small-effect interpretations
this paradigm should not be used as a manipulation check
two failures undermine reliability for this use, but independent laboratories were not tested
Derived from the full evaluation — not a separate score.
Strengths
The discussion calculates that only 14 participants would be needed for the original d=2.29 effect and that Study 1 could detect a conventional large effect, directly supporting rejection of a large effect.
↳ Discussion, paragraph beginning “In the case of the envelope task”
Both studies randomly assigned participants to oxytocin or placebo and used colleague-held allocation codes to maintain double blinding until study completion.
↳ Study 1 Participants; Study 2 Participants
The discussion evaluates five explanations spanning false positivity, experimenter effects, small effects, reversed effects, and pharmacokinetic uncertainty rather than treating non-significance as proof of no effect.
↳ Discussion, five-hypothesis analysis
Limitations
Participants were informed about possible adverse effects and allergic reactions, but neither Results section reports systematic adverse-event collection or outcomes after intranasal oxytocin.
↳ Ethics Statement; Study 1 Results; Study 2 Results
Study 1 used a different questionnaire and is acknowledged as conceptual, while both studies came from one laboratory and included only healthy young adult males.
↳ Discussion opening; Study 1 Participants; Study 2 Participants
The evidence directly addresses the Envelope Task under closely related laboratory conditions and does not empirically establish generalization to other trust paradigms, laboratories, or populations.
↳ Study 1 and Study 2 Methods; Discussion opening
The central replication claim is supported by consistent null Envelope Task results and power calculations directed at the original d=2.29 effect. Randomization, double blinding, baseline checks, and convergent ANOVA and ordinal-regression analyses support competent execution, while absent adverse-event results reduce the Methodological Rigour assessment. The abstract and five-hypothesis discussion calibrate the main conclusion to rejection of a large effect rather than proof of no effect. Impact remains primarily methodological because both studies use healthy young adult males in one laboratory and address a single paradigm.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Strong3.8
Two failed replications materially constrain a previously reported very large effect while preserving uncertainty between a null and a small effect. The advance is meaningful but restricted to one trust paradigm and research group.
“clearly exclude a large effect of OT on trust in this paradigm”
Both studies use random assignment, double-blind administration, baseline comparisons, and ANOVA plus ordinal regression. No systematic adverse-event results are reported despite administering intranasal oxytocin, which lowers the score within the competent-methods band.
“The administration followed a double-blind procedure”
The paper moves coherently from the original effect through two null studies to five possible explanations and clearly distinguishes exclusion of a large effect from uncertainty about small effects. The studies’ separate parent aims and some conditional field-wide speculation add structural and interpretive friction.
“we do not know which hypothesis is true”
The discussion engages failed replications, meta-analytic skepticism, publication bias, pharmacokinetic uncertainty, and potential moderators. Limitations generally alter the stated scope of interpretation, although Study 1’s conceptual-replication status receives less detailed treatment than other alternatives.
“Study 1 is not a direct replication of the original study”
Caveats4 of 4 checks
The principal Envelope Task results are numerically coherent and consistently reported, but one secondary positive result is contingent on outlier removal and the power claim requires careful attention to effect-size scope.
Ethics approval, informed consent, random assignment, and double-blind administration are declared. Registration, data availability, code availability, and conflict disclosures are not established by the supplied text but do not constitute demonstrated conduct failures.
Flags: 1 declared / 5 total
35 of 35 checkable references verified
36 references in manuscript 1 are books, websites or datasets — counted, but not index-checkable
No retraction notice found in Retraction Watch.
Sources: Retraction Watch ✓
Medium3.0
The paper gives oxytocin researchers, reviewers, and meta-analysts a concrete methodological decision: whether to continue using the Envelope Task as a manipulation check. Its audience is identifiable but predominantly academic.
“this paradigm should not be used as a manipulation check”
The findings can change laboratory design and interpretation but do not establish a clinical intervention, policy instrument, or deployment-ready practice. Readiness therefore consists chiefly of research-course correction.
“OT literature should be more open to non-significant results”
Consistency across two studies supports transfer to closely similar Envelope Task settings. Both studies nevertheless involve healthy young adult males from one laboratory using the same dose and related procedures.
“healthy young adult males”
The replications are explicitly situated alongside prior failures, meta-analytic estimates, and calls for preregistration-style discipline and fuller reporting. This gives the result a credible role in cumulative correction of the oxytocin literature.
“Replication studies add great value to science”
AI-generated, human-governed. Something look off? Contact us to request a review.