Related Experiment Video
Updated: Apr 8, 2026

Measuring the Subjective Value of Risky and Ambiguous Options using Experimental Economics and Functional MRI Methods
Published on: September 19, 2012
Exact optimum coin bias in Efron's randomization procedure
Alessandro Baldi Antognini1, William F Rosenberger2, Yang Wang2
1Department of Statistical Sciences, University of Bologna, Via Belle Arti 41, 40127, Bologna, Italy.
We introduce a new method to optimize Efron's biased coin design, balancing treatment assignment and randomization. This approach enhances clinical trial efficiency by selecting the optimal bias parameter (p) for improved study outcomes.
Area of Science:
- Biostatistics
- Clinical Trial Design
- Randomization Methods
Background:
- Efron's biased coin design is a restricted randomization procedure offering favorable balancing properties.
- It maintains full randomization, with treatment assignment probabilities less than 1.
Purpose of the Study:
- To propose a compound optimization strategy for Efron's biased coin design.
- To determine the optimal bias parameter (p) by weighting balance and randomness criteria.
Main Methods:
- Utilizing exact and asymptotic distributional properties of Efron's coin.
- Developing a compound optimization strategy based on subjective weighting of criteria.
- Analyzing imbalance variability, expected imbalance, selection bias, and accidental bias.
Main Results:
- The proposed strategy identifies an optimal bias parameter (p) for Efron's coin.
- The method is applicable to both small/moderate and large sample clinical trials.
- It balances the trade-offs between allocation balance and randomization probability.
Conclusions:
- The compound optimization strategy provides a robust method for selecting the bias in Efron's design.
- This enhances the efficiency and reliability of restricted randomization in clinical trials.
- The findings are relevant for optimizing statistical power and minimizing bias in comparative studies.
More Related Videos
Related Concept Videos
Bias
In statistics, a sampling bias is created when a sample is collected from a population, and some members of the population are not as likely to be chosen as others (remember, each member...
Randomized Experiments
Simple randomization
Simple...
Bonferroni Test
The means of different samples are first paired in all possible combinations.
The null hypothesis of the...
Friedman Two-way Analysis of Variance by Ranks
Expected Frequencies in Goodness-of-Fit Tests
Fisher's Exact Test

