"Selfield" immersive environment precognition study

A preprint reporting a forced choice precognition study using an "immersive audiovisual environment to induce a psi-conducive state", comparing results with and without feedback, and looking at the effects of mental discipline/meditation, has been uploaded to ResearchGate:
https://www.researchgate.net/publication...n_Research

There were four pre-registered hypotheses, and the results were non-significant at 5% (though in the expected direction) for all of them.

The Selfield: Optimizing Precognition Research

Mario Varvoglis, Peter A. Bancel, Jean-Paul Bailly, Jocelyne Boban and Djohar si Ahmed

We report an exploratory forced-choice precognition study based on a protocol that utilized an immersive audiovisual environment to induce a psi-conducive state in participants. Our objective was to assess whether this optimization setup would help produce significant psi results with an unselected population. We also sought to assess whether trial-by-trial feedback would produce superior scoring to no-feedback trials. For each trial, participants selected an opaque graphical sphere that they felt contained a facial image (as opposed to being empty). After selection, the program randomly determined whether the sphere would be empty or not, and whether feedback should be shown. A preset total of 3000 binary choice trials were collected from 82 participants. Each participant contributed either 1 or 2 20-trial series, based on preset scoring criteria. The total hit rate of successful trials was 50.1%, close to expectation under the null hypothesis of no psi effect. Hit rates for feedback and no-feedback trials were in the predicted direction (51.0% vs. 48.6%). A post-hoc analysis showed that hit rates for feedback trials increased over the 20-trial series, suggesting that participants may have progressively found a mental strategy for improved scoring. Additionally, a subgroup of 26 experienced meditators had a hit rate of 52.1%, a result consistent with previous literature that suggests that meditators are particularly good participants for psi research. 

I've now read the paper, which gives a clearer idea of the motivation behind the work, which is essentially to try to improve the effect size in forced-choice experiments (i.e. experiments in which the subject's response is to choose between a fixed set of alternatives). The effect size is generally believed by parapsychologists to be larger in free-response experiments (i.e. experiments where the subject has more freedom in expressing his/her response - albeit that response will still ultimately be applied to a fixed set of alternatives), but these require much greater resources than forced-choice experiments.

This study follows an earlier one along the same lines by (almost) the same authors published in 2013, which was based on telepathy between two subjects and included various features (including an "immersive" environment) intended to optimise the psi effect. That study also produced results which were statistically non-significant overall, but in which post hoc analysis showed more variability when the optimisation features were present than when they were absent, suggesting that they might be having an effect, though not a consistent one. A rather detailed abstract of the earlier study is at p. 181 here:
https://www.rhine.org/images/jp/JPv77n2.pdf

So the new study modified the earlier one in several ways, including improving the immersive technology (which some subjects had found distracting), making the targets more arousing, and testing whether feedback about the correctness of choices was necessary (because some had also found the feedback distracting). The protocol was also simplified by replacing two-subject telepathy with single-subject precognition.

But again there was a non-significant overall result, so it seems the effort to raise the forced-choice effect size towards free-response levels was unsuccessful. From the direction of the differences, and from some secondary analyses which were statistically significant, the conclusion was that the results are better with feedback (perhaps not surprisingly in a precognition experiment) and for subjects with experience of meditation and good hypnotic participants - particularly a group of meditators in the Shambhala lineage of Tibetan Buddhism. So the authors recommend an "elitist" approach for future studies, in which participants with experience of these mental disciplines are sought.

Perhaps it's worth noting that the study used a pseudo-random number generation, based on the timing of two mouse clicks by the participant - the first before the trial, which generates a sequence of bits, and the second at the point when the choice is made, which determines which of the bits is used. I can't see any discussion in the paper of why this method is used.