It is a very important aspect of the scientific method that any experiments be, in principle, replicable by other people. No matter what you are investigating, in order for your research to be scientific, some other person must be able to do the same experiments to see if she comes up with the same results.
In addition, it must be possible to introduce additional “controls” to the experiment in order to determine what exactly is happening. For example, if you have performed an experiment which you think demonstrates that alcohol produces certain health benefits, another person might introduce the additional control of only using wine in order to see if any form of alcohol shows such benefits or if only certain drinks, like wine, are good for you.
R.A. Fisher, in this book The Design of Experiments, writes:
- “In order to assert that a natural phenomenon is experimentally demonstrable we need, not an isolated record, but a reliable method of procedure. In relation to the test of significance we may say that a phenomenon is experimentally demonstrable when we know how to conduct an experiment which will rarely fail to give us a statistically significant result.”
A single experiment which provides us with anomalous data, even if statistically significant, does not provide us with a valid justification to assert that we have demonstrated the existence of a particular natural phenomenon as the cause. Instead, multiple experiments performed by multiple people, all replicated in varying ways, must be available first. According to Susan Blackmore:
- “One of the arguments against the evidence for psi has always been that only the good results are published. If a hundred experiments are carried out, you will expect — by sheer chance — that about five will give results “significant” at the .05 level. If only those five are published, and the other ninety-five forgotten, the published evidence will be quite unfairly biased. It seemed feasible that this selective reporting accounted for much of the significance reported in the literature.”
In the long run of many, many experiments, the chances are high that a few will produce anomalous results — i.e., results which are not expected by sheer chance. These results seem important because people only look at them in the context of the short-run, namely that in a single experiment such results should not occur. David Marks explains such misunderstanding thus:
- “It is a simple deduction from probability theory that an event that is very improbable in the short run of observations becomes, nevertheless, highly probable somewhere in a long run of observations. For example, if we flipped five coins at once, the probability of getting five heads is 1/32, or about .03. But if we repeated the flipping of five coins ten times, the probability of getting five heads somewhere in the ten tests is about .27. If we ran 100 tests, the probability of five heads rises to .96, which is highly probable indeed. [a probability of 1.0 is a certainty] But if we stopped anywhere in these 100 tests and asked what the probability would be of getting five heads on the very next trial, we are back to the starting probability of .03 because we have switched from a long-run question to a short-run question.”
That is why scientific experiments require repeatability — a single anomalous result in a long string of expected results is not sufficient cause to rethink all of our understandings of the universe. In the field of psychic research, however, we have exactly the opposite situation. One or two experiments are performed on the same person by the same researchers, often with very poor controls and standards. When others attempt to replicate this experiment and/or when more stringent controls and standards are introduced, suddenly the alleged phenomenon disappears — but that is ignored and parapsychologists claim to have evidence that the laws of physics are much different than anyone realizes.
If parapsychology were a genuine science, why has none of its work become incorporated into other science? Although every field in science is distinct, they are also connected by the fact that the phenomena they study rest upon the same biological, chemical and physical laws. Thus, when some unexpected phenomenon in chemistry is noted, possible biological explanations are offered; when some unexpected phenomenon in biology is noted, possible chemical explanations are offered.
Yet, none of these scientists will ever suggest ESP or telekinesis as possible explanations for unexpected or unusual phenomena. Indeed, none of them would even think to make such a suggestion. For the scientists, those would be completely useless explanations — they would not make the phenomena any more understandable, nor provide any testable predictions, and they would not move the research forward one iota.
