Minimization is a method whose goal is to minimize imbalance of subject characteritics between study arms. This method is best explained with an example. Suppose we have a study in which we are interested in minimizing sex and race imbalances between the two study arms, treatment and control. Thus far, we have enrolled 36 subject. These 36 subjects have the are summarized by characteristic in the neighboring table.

Suppose we are enrolling a Hispanic female as our new subject. To allocate her according to the Minimization algorithm, we take the number of Hispanics in the treatment arm and adds it to the number of females in the treatment arm. This value is 12. Then we add the number of Hispanics in the control arm and add it to the number of females in the control arm. This value is 11. We then allocate her to the study arm with the smaller value, in this case the control arm.

Note: if a characterisitc is continuous, such as age, it needs to be binned (e.g. '< 18', '18-35', '36+').

Characteristic Level Treatment Control
Sex: Female 10 9
Male 8 9
Race: African American 3 4
Asian 2 0
Hispanic 2 2
White 11 12

List Details

Sign Up to Begin List!


{{ this.covariate_errors }}

No characteristics have been added yet.
Characteristic Name Levels
{{ }} {{ covariate.levels }}

New Characteristic

Sign Up to Begin List!


  • Taves, D. R. (1974), "Minimization: A new Method of Assigning Patient to Treatment and Control Groups," Clinical Pharmacology and Therapeutics, 15(5), 443-453.
  • Scott, M.A, McPhearson, G.C, Ramsay, C.R, & Campbell, M.K. (2002), "The method of minimization for allocation to clinical trials: a review," Controlled Clinical Trials, 23, 662-674.