Bias Testing: Bias Test Visualization Using Pierre Gy's Approach

The approach described in this chapter was introduced by Pierre Gy to help with the interpretation of the results of bias tests. It makes use of the Student’s t -distribution to evaluate whether the observed bias for an estimator is statistically significant, but in contrast to the usual bias test it repeats the evaluations after each trial and plots the results on a control chart. The advantage of this approach is to provide practitioners performing bias tests with additional insights into what may be influencing the overall test results. Providing answers to questions such as: Are the results of the trials trending in a particular direction? If they are trending, how many more trials are needed to change the outcome of the bias test? Are there any anomalous trials that significantly influence the outcome of the test?