JonTaylor

I notice that the duration of the runs was randomly selected in your pilot experiment. However, I couldn't determine whether the moment at which each run commenced was also randomly determined or dependent on a human operator. After several years working on the subject of intuition, I am convinced by Ed May's decision augmentation theory (DAT) according to which a participant uses precognition (in the form of intuition) to choose the moment at which to generate a data sample, in such a way that it conforms to the intended result. For example, in the REG experiments, the operator of the REG uses intuition to sample the bit stream at a moment for which the sample will later be found to contain the intended deviation in the mean value for the output distribution. Thus, if the result of your experiment could be affected according to the moment at which the runs are started, such human control could introduce an artifact. In this case, it might be worth automating such control, or better still, comparing the effect of automatic Vs manual control.