I.05 / GNOSIS
Meta-science
Which findings survive replication, and which features of the original study account for the difference?
Start with the question.
Meta-science studies the machinery by which findings are produced, checked, circulated, and remembered. The Open Science Collaboration's coordinated psychology project offers a documented example of examining replication across a selected group of studies. Its value here is methodological: the unit of analysis becomes a research process, including sampling, analysis choices, access to materials, and the criteria used to call two results consistent.
Inspect a finding's route into a summary.
Create a table for five papers on one question. Record the planned outcome, reported outcome, sample size, access to materials, and whether an independent follow-up exists. Keep an absent record as missing information rather than silently treating it as a failed experiment.
What would you inspect?
- The denominator used for any reported success or replication rate.
- Differences between an effect estimate, its uncertainty, and a significance threshold.
Take the next step.
Select one field-specific claim about reliability and reconstruct the studies included and excluded before comparing its reported rate with another survey.
Read the supporting material.
- Estimating the reproducibility of psychological science
Open Science Collaboration / Science · A coordinated replication project and its stated design; supports study-level reproducibility questions.
