Calculators
Use the browser calculators directly for prevalence, elusion, recall, richness, precision, control-set recall, and culling-impact questions.
Enter values in any card below to calculate live outputs. The hosted MCP exposes related tool-callable versions for agent workflows.
Calculator
Estimate prevalence, confidence bounds, responsive volume, and review burden.
Prevalence = r / n; estimated responsive = N x prevalence; burden range = N x CI bounds x burden multipliers.
Use a random sample from the review population to estimate likely responsive volume before review planning.
Enter values to see calculator output.
Calculator
Estimate elusion rate, remaining responsive material, and recall bounds from an elusion sample.
Elusion rate = responsive in elusion sample / elusion sample size; recall = responsive found / (responsive found + estimated remaining responsive).
Use after ranking or review to estimate what may remain in the unreviewed or discarded population.
Enter values to see calculator output.
Calculator
Convert a sample into richness, estimated responsive volume, and a yield proxy.
Richness = r / n; estimated responsive = N x richness; yield proxy = richness.
Use early samples to understand expected responsiveness and communicate likely review yield.
Enter values to see calculator output.
Calculator
Calculate precision, recall, and F1 from a confusion-matrix slice.
Precision = TP / (TP + FP); recall = TP / (TP + FN); F1 = 2 x precision x recall / (precision + recall).
Use adjudicated classifications to compare model or workflow performance.
Enter values to see calculator output.
Calculator
Validate recall against the responsive portion of a control set.
Observed recall = responsive found in control / responsive in control; responsive shortfall = responsive in control - responsive found.
Use a control set when you need a direct recall check against known responsive documents.
Enter values to see calculator output.
Calculator
Compare non-responsive removal against responsive loss to quantify selectivity.
Culling efficiency = non-responsive removed percentage / responsive removed percentage; responsive retention = R1 / R0.
Use before-and-after counts to check whether culling removes noise without disproportionate responsive loss.
Enter values to see calculator output.
Use note
Results are planning aids and can be wrong if inputs or assumptions are wrong. Users remain responsible for independent human and professional review.