Open Experiment Standard
Section 6 · OES Spec

Analysis configuration

How the scorecard was computed — reproducibility starts here.

A result is not 'conversion +3%'. It is 'conversion +3% under this model, window, population, and variance estimator'. The analysis block makes those choices explicit and machine-readable.

Fields

FieldTypePurpose
analysis.methodenumfrequentist, bayesian, sequential, cuped, diff_in_diff, etc.
analysis.modelstringt-test, regression, bootstrap, Bayesian normal, binomial, etc.
analysis.varianceEstimatorenumnaive, delta_method, cluster_robust, sandwich, bootstrap.
analysis.confidenceLevelnumber (0–1)Usually 0.95.
analysis.alphanumber (0–1)Significance threshold.
analysis.priorobjectBayesian prior, if relevant.
analysis.adjustmentMethodstringCUPED, regression adjustment, stratification, etc.
analysis.multipleComparisonCorrectionstringCorrection method.
analysis.segmentationstring[]Segments included in analysis.
analysis.dimensionBreakdownsstring[]Browser, country, new/returning, plan type, etc.
analysis.missingDataHandlingstringHow missing outcomes were treated.
analysis.outlierHandlingstringWinsorization, trimming, capping.
analysis.queryReferencesobject[]SQL query IDs, warehouse job IDs, notebooks.
analysis.generatedAtstring (date-time)Time the result was generated.