Community
RPP vs Many Labs
Cross-Dataset Comparison

RPP vs Many Labs:
What Two Replication Projects Agree On

The same contradiction detector applied to two independent large-scale replication projects. Where both agree, the finding is the most reliable in this analysis.

KEY FINDING

The contradiction rate is similar across both datasets (RPP: 36.3%, Many Labs: 30.8%), suggesting this is a stable property of the literature rather than an artifact of either replication project's methodology. The flag priming effect is the only direct contradiction confirmed by both independent replication projects — making it the highest-confidence contradiction in this analysis.

168 STUDIES
OSF Reproducibility Project: Psychology (RPP)
18
Direct
43
Partial
39
Replicated
36.3%
Contradiction Rate

40.5% of studies have missing direction data, concentrated in incomplete replications.

13 STUDIES
Many Labs 1 Replication Project
2
Direct
2
Partial
9
Replicated
30.8%
Contradiction Rate

All 13 studies have complete direction data. Many Labs used 36 sites and 6,000+ participants per effect.

RPP Distribution
Direct
18
Partial
43
No Contradiction
39
Insufficient Data
68
Many Labs Distribution
Direct
2
Partial
2
No Contradiction
9
Insufficient Data
0

Cross-Dataset Overlap

CONFIRMED BY BOTH DATASETS

Flag priming and political attitudes (Carter et al. 2011) appears as a direct contradiction in BOTH datasets — RPP Study 9 and Many Labs ML09. This is the strongest finding: a direct contradiction confirmed independently by two large-scale replication projects.

RPP-ONLY CONTRADICTIONS

Ego depletion, power posing, and behavioral priming effects appear as direct contradictions in RPP but were not tested in Many Labs 1.

MANY LABS-ONLY CONTRADICTIONS

Low-effort thought promoting conservatism (Eidelman et al. 2012) is a direct contradiction in Many Labs but was not in the RPP dataset.

Direct Contradictions by Dataset

RPP — 18 direct contradictions (sample)
RPP-03

Ego depletion effect

Depleted participants perform worse on subsequent self-control tasks

RPP-04

Priming effect on behavior

Priming with elderly concepts slows walking speed

RPP-05

Power posing effect

High-power poses increase risk tolerance and testosterone

RPP-09

Flag priming and political attitudes

Exposure to American flag shifts attitudes toward Republican positions

RPP-56

Anger in bargaining

Angry recipients receive lower offers when consequences of rejection are low

RPP-61

Poignancy and mixed emotions

Graduates reminded of meaningful endings report more mixed emotions

RPP-77

Goal framing and motivation

"To-date" vs "to-go" framing affects donation behavior

RPP-107

Nonconscious goal pursuit

Goal primed participants learned more than control group

RPP-118

Ego depletion and attraction effect

Depletion with placebo increases attraction effect

MANY LABS — 2 direct contradictions
ML04confirmed

Low-effort thought promotes conservatism

Cognitive load increases conservative attitudes

ML09confirmed

Flag priming and political attitudes

Exposure to American flag shifts attitudes toward Republican positions

9 SUCCESSFULLY REPLICATED

Anchoring, reciprocity, cognitive dissonance, mere exposure, sunk cost, and others replicated across 36 sites.

RUN YOUR OWN DATASET

Upload any CSV containing replication study data to the contradiction detector. The tool runs client-side — your data never leaves your browser.

Open the Contradiction Detector →

AI ASSISTED RESEARCHER · MANUS AI · RPP vs MANY LABS COMPARISON · MAY 2026
Last updated: May 2026
RPP data: Open Science Collaboration (2015). Many Labs data: Klein et al. (2014).