Toggle Menu

Tool Instructions

Creating A New Trial

To create a new randomized trial, go to the Trial Randomization Tool. After entering your trial's parameters and clicking "Send Confirmation Email," you will receive an email with further instructions on how to download your randomized results.

Please note that we limit users to a maximum of three trials queued for randomization at a time. If you wish to submit additional trials, you must wait until the output from one of your earlier trials has been generated.

The input parameters for this tool are described below:

Parameters

The word "subject" is used here to mean a participant in your trial.

  • Trial Name:

    A short identifier for your trial. The download file containing your results will be labeled with this name. Emails that we send you pertaining to your data will also mention your trial with this name.

  • Number of Trial Arms:

    The number of arms in this trial. For each arm, you must provide a name. Common examples might be "Treatment" and "Control."

  • Arm Allocation Ratio:

    The ratio used to allocate subjects into trial arms. Each arm can be assigned its own ratio relative to the first arm.

  • MTI:

    The "maximum-tolerated imbalance." For a trial with a 1:1 arm allocation ratio, at any point in the randomization process, no trial arm will be assigned this many subjects more than any other arm.

    For arm allocation ratios other than 1:1, the imbalance between two arms is calculated as I = | R1 * C2 - R2 * C1 | where I is the imbalance, R is the allocation ratio for a given arm, and C is the number of allocations for a given arm. At no point do we allow this imbalance to deviate beyond the number prescribed by the maximum tolerated imbalance.

  • Trial Size:

    The number of subjects in your trial.

For Stratified Trials

You may select whether or not to stratify your trial. If so, additional inputs are provided:

  • Stratum Size:

    The number of subjects in the largest stratum of your trial. This many random assignments will be generated for each stratum. If you are uncertain how many subjects will be in the largest stratum, it is safe to generate more assignments than what ends up being necessary. It is assumed that more than the necessary number of assignments will be generated for smaller strata.

  • Number of Stratified Variables:

    The number of variables you wish to stratify on. For each stratum, you must enter a name, and the number of possible values. You will then be able to specify what those values are.

    A different randomized allocation sequence will be generated for every combination of the possible values of these variables. For example, you have two stratification variables with three possible values each, six sequences will be generated for each of the six combinations. Common examples for these variables include gender, race, age, and testing center.

Brief descriptions of these fields can be accessed by clicking the question marks on the Randomization Tool form.