Two Attempted Retro-Priming Replications Show Theory-Relevant Anomalous Connectivity

We report a theory-relevant post hoc analysis of two Dutch retro-priming experiments that were part of a large replication project of the retro-priming experiment by Daryl Bem and colleagues. This replication project sought to investigate the role of the experimenter in psi studies. The results of the retro-priming experiments performed by student research groups at the University of Amsterdam (N = 61) and the University of Groningen (N = 222) did not, however, replicate Bem’s earlier findings of an anomalous interference of a future stimulus on response times. We report the results of these two studies here, but the overall results will be reported elsewhere. Both Dutch studies used the exact same software as did Bem and colleagues. However, each study used a different questionnaire. The questionnaires asked for information that in previous research had been associated with success in psi tasks and that could help us to deal with individual differences, but above all could be used as selection criteria for participants in future studies. In the Amsterdam study, there were 14 questions, while in the Groningen study there were 55. A correlation analysis revealed several significant correlations between the psi effect in the Bem task and questionnaire items. In this paper we focus on the post-hoc research question: Is this global composition of the correlation matrix anomalous, as suggested by Generalized Quantum Theory? Rather than using the subjective number of ‘significant’ correlations as a dependent variable, we introduced two objective measures directly representing the correlation values in the cells to characterize the ‘Connectivity’ in the matrix. Our analysis revealed ‘Connectivity’ to be marginally significantly larger (p < 0.075) in the Groningen study and significantly larger in the Amsterdam study (p < 0.025). These results are discussed in the framework of the Consciousness Induced Restoration of Time Symmetry (CIRTS) theory and the Generalized Quantum Theory (GQT) that predict that as soon as there are reasons to expect replicability (or control) the extra chance psi scores will disappear or 44 J a c o b J o l i j a n d D i c k B i e r m a n reverse. The conclusion is that these results can be seen as support, though statistically not strong, for the GQT. Introduction In spite of strong experimental claims by some proponents, for instance Bem et al. (2015), of the psi hypothesis, the reality of claimed phenomena such as telepathy has not been accepted by mainstream science (e.g., Traxler et al. 2012). The major reason given for this skeptical position is that there appears to be no parapsychological experimental paradigm yet with a robust effect size that might be used by independent researchers to check this reality for themselves, i.e. given a specific power and a specific sample size, a specific distribution of outcomes as well as the probability for success in a single experiment can be calculated. Thus independent researchers can check if their experiments do follow the distribution of outcomes as predicted by the power of the original studies. This problem of nonreplicability is characteristic in controversial research areas, but during the last decade has also been a topic of intensive discussion in mainstream science because in a number of ‘well-established’ fields it turns out that the replicability rates are suspiciously low (Pashler & Wagenmakers 2012). Within the field of parapsychology there are researchers who rather consistently seem to get results confirming the psi hypothesis, and there are independent researchers who fail to do so, and for these researchers improvement of the paradigm to a level where a specific power can be expected is high on the wish list. The Heymansgroup of the University of Groningen, embedded in the Department of Experimental Psychology, developed a multi-year research program to that end (Heymansgroup.nl). If replication is such a problem, there has to be a large portion of unexplained and uncontrolled variance. This uncontrolled variance may be due to so-called questionable research practices or QRPs (Bierman, Spottiswoode, & Bijl 2016). However, as a result of further automation, especially the real time uploading of experimental data (Jolij & Bierman 2017), as now is becoming standard at the Heymansgroup and other research groups worldwide, it seems that these known QRPs are becoming increasingly unlikely. The other obvious source for this error variance could be individual differences in participants. However, as is often claimed in experimental parapsychology, the uncontrolled variance may also be due to differences in experimenters. Therefore, the Heymansgroup embarked on the development of a selection instrument to be used in future experiments for participants as well as experimenters (see mission statement at Heymansgroup.nl). This selection instrument would then become publicly available in order to give R e t r o Pr i m i n g R e p l i ca t i o n s S h o w A n o m a l o u s C o n n e c t i v i t y 45 independent replicators the tools to reduce error variance due to individual differences in participants and experimenters. To begin with, the group decided to use a coarse-grained approach to explore simple questions to help predict performance in a parapsychological experiment that was run in another context as sketched above. Student experimenters had to first study the relevant literature searching for predictors of psi and come up with questions that captured the suspected predictor. The historical search for personality variables predictive of psi has had limited success. One of the successful and more consistent predictors of subjects’ performance is ‘belief in the reality of psi’ (Schmeidler & McConnell 1958). However, even here the results seem ambiguous, often caused by nonbelievers scoring ‘the wrong way’. Other predictors that have positively been implicated are ‘experience with mental disciplines’, ‘creativity’, ‘intuitive versus rational thinking’ (Honorton 1997). Thus the questionnaires in both studies contained a number of questions relating to these potential predictors (see Appendix 1 for the wording of these questions). In spite of this study being highly explorative with regard to the questions intended to measure factors that had been associated with positive scoring in a psi task in the past, the student experimenters expected specific positive correlations in the literature they had been studying. In the following we will describe the experiment in Groningen in detail in the Methods section. The experiment in Amsterdam that was done one year earlier was identical but used 14 different questions to be correlated with psi performance.