Works pseudo-telekinesis

Psi research and skepticism

Psi is the power that is said to be behind fortune telling, telepathy and telekinesis and many other wonders. The psi topic is a long-running issue in the skeptic movement and it also boils up again and again in the private sector.

In the second half of the last century, the psi was the subject of serious research. In the meantime, scientific interest in it has waned. But that does not reduce the popularity of the Psi.

If you want to get to know the scope of scientific methods, this area is an excellent playground. The opponents in the field are on the one hand the psi researchers and on the other hand the activists of the skeptic movement.

The psi researcher assumes that psi exists. If necessary, he can be taught better. The psi skeptic assumes that psi does not exist. If necessary, he can be taught better.

The Ganzfeld data

One result of Psi's scientific investigation is that Ganzfeld data set (Ganzfeld Database). It comprises the results of 128 studies from 1974 to 2018 that follow the so-called Ganzfeld Protocol. The aim of this protocol is to rule out deceptions and self-deception as far as possible (Broderick, Goertzel, 2015, Through Time and Space, p. 168 ff.).

In these studies, each elementary test tests the psi ability of a subject (recipient) by identifying one of four possible targets that another person, the sender, “keeps in mind”. The target and the sender are hidden from the recipient's senses. He only knows the four possibilities. To identify the chosen target, he only has the psi.

This test arrangement and the associated statistics correspond to those of the Würzburg Psi-Test.

Under the null hypothesis (no psi), the hit probability in the Ganzfeld test is 25% per run. This elementary test is carried out several times in each of the studies recorded in the database. We know the number of each of the hundred or so studies N these elementary tests and the hit rate. The latter is equal to the total number of hits in the study divided by N (relative hit frequency).

Under the null hypothesis, i.e. under the assumption that Psi does not work, the number of hits per study of the (N, 25%) - binomial distribution. The limit values ​​for significance levels can be determined under this condition.

In the funnel graph below, the funnels are determined by the limit values ​​for the 5% and 1% significance level. Within the funnel, 95% or 99% of the hit rates are expected under the condition that the null hypothesis applies. Values ​​outside a funnel are significantly at the appropriate level. The hit rate and the sample size are entered in the funnel graphic for each recorded study N.


The Ganzfeld dataset has been extensively analyzed and a number of meta-analyzes have been carried out. The rules for the meta-analyzes (selection and weighting of the individual studies) generally differ from researcher personality to researcher personality. It is therefore no wonder that some people come to the conclusion that the data shows nothing special and another, in turn, reaches an incredibly high level of significance in favor of Psi.

In the analysis of the whole field data, the psi researcher Charles Honorton excelled and the skeptic Ray Hyman.

Anyone who wants to present the status of the discussion between psi researchers and skeptics as objectively as possible must also take into account the point of view of the other side. I am listing here what seems to be indisputable between the representatives of psi research and the skeptics. To this end, I, the skeptic, deliberately consult the work of the psi researchers. They are given as sources.

  1. Psi effects are, if they exist, weak and moody;Scientific tests in the classical manner reach their limits.
  2. Metastudies do not help much. The researcher's preferences can have an impact on the analysis result. The selection and weighting of the studies depend on the individual judgment on suspected cases questionable research practices like Fishing for Significance (Bierman et al., 2016).
  3. The ones underlying the possible psi effects causal mechanisms are still unknown.
  4. A practical use the psi research is not in sight. The major sponsors of Psi research, the US military and Sony, have suspended their related programs.

Controversial: the yardstick

The skeptics are accused of continuing to raise the bar of significance after all the efforts and successes of the psi researchers. When the skeptic Carl Sagan says, "Extraordinary claims require exceptionally strong evidence," then the psi people can already feel touched.

As a general hint, this skeptical motto may still pass; strictly speaking, there is something anti-progressive about it. The skeptic gets completely on the wrong track if he wants to elevate this motto to a law and uses the Bayesian formula to do so, i.e. if he thinks that proof of psi cannot lead very far in view of the low a priori probability.

Such arguments lead nowhere; namely, there is no generally accepted definition for the hypothesis probability. The frequency interpretation also offers no way out: Bayesian estimates are misleading in this context.

On the other hand, the skeptic is absolutely right when he accuses the psi researcher of lowering the bar of effect size so that even the weakest psi could overcome it. A weak and moody psi cannot be identified with statistical methods and certainly cannot be ruled out.

What is the purpose of psi tests today?

The first point above shows that the Würzburg Psi-Test, for example, will be unsuccessful. Neither positive results nor convincingly negative results are to be expected. The psi is - if it exists - obviously too weak and moody for that. Since the tests follow uniform rules, the meta-study of all tests carried out so far can be carried out without any problems. But even this summary of the tests has so far not revealed anything special. And that is not to be expected in the future either.

The Würzburg Psi-Test still makes sense. It is less about the question of whether there are psi and also not about whether one should continue to struggle with the evidence for or against psi in view of the evidence. The Psi-Test shows its value in educational work and in the demonstration of scientific working methods. The four points listed above should be at the center of this educational effort.


Bierman, Dick J .; Spottiswoode, James P .; Bijl, Aron: Testing for Questionable Research Practices in a Meta-Analysis: An Example from Experimental Parapsychologie. PLOS ONE, 2016

Broderick, Damien; Goertzel, Ben (Ed.): Evidence for Psi. Thirteen Empirical Research Reports. 2015

This entry was filed under Education, Science, School and College, Skepticism, Probability Theory, Science and Pseudoscience and tagged Enlightenment, Falsifiability, Parapsychology, Significance, Statistics, Sampling, Deception and Self-Deception, Probability Calculation. Set a Bookmark the permalink.