Research Workbench

Minimization imbalance comparison

Compare weighted marginal imbalance for the next synthetic allocation.

Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.

Existing A count, existing B count, factor weight; matching level for each factor

Enter the counts for the new subject’s level of each balancing factor, not all levels. Counts are nonnegative integers and weights positive; use 1–20 factors.

The starting values are hypothetical. Results update as you edit. Use aggregate or synthetic inputs only.

Calculation results

If assigned A: weighted imbalance: 5
If assigned B: weighted imbalance: 4
Balancing preference: B has lower marginal imbalance.

How to interpret the output

For each candidate allocation, the score sums weight×absolute(A−B) after adding one to that arm. The lower score is a balancing preference, not a randomized assignment. An operational minimization procedure needs a prespecified random component, concealment, governance and simulation; this tool does not allocate participants.

Method reference

Source checked 2026-11-06. Read the reproducible calculator validation report.