These are two important factors of genuine science: controls and repeatability. Scientific theories are based upon and lead to controlled, repeatable experiments whereas pseudoscientific theories are based upon and lead to experiments which are not controlled and/or are not repeatable.
Controls means that it is possible, both in theory and in practice, to eliminate possible factors which might be affecting the results. As more and more possible factors are cut out, it is easier to claim that only one particular thing is the "real" cause of what we see.
Repeatability means that we are not the only ones who arrive at our results. In principle, it must be possible for other, independent researchers to perform the exact same experiment and arrive at the exact same results. It is not enough for "believers" to repeat the experiments - phrenologists managed to repeat each other's experiments in the 19th century, but phrenology did not turn out to be valid.
Peers & Science
When our experiments are successfully replicated by others, our theory and our results are further confirmed. This may not be quite enough to call it "proven," but given enough time and further replications that will be the likely conclusion. If our experiments are unsuccessfully replicated, especially if more than once, then we have good reason to revisit our claims and reconsider our results.
In parapsychology neither controls nor repeatability appear to be common. Controls, when they exist at all, are typically very lax. The scientists who perform the experiments are rarely if ever prepared to deal with the possibility that a subject might wish to deliberately trick them. When controls are sufficiently tightened to pass regular scientific scrutiny, it is common that the paranormal results no longer appear.
When scientific standards have managed to make some headway, there is often the excuse that the presence of "skeptics" creates "negative vibrations" which conveniently prevents the phenomena from appearing. As a result, carefully constructed controls tend to be abandoned in order to appease the claimants of psychic powers - and of course, the powers suddenly reappear.
Repeatability is also something in short supply in the field of parapsychology. This tends to be called an "experimenter effect" - some experimenters are able to elicit "real" psychic phenomena whereas others, quite inexplicably, cannot. This runs contrary to the whole of professional science which relies upon the idea that objective, independent observation and verification are possible. According to Susan Blackmore:
"In most other fields there is rarely a conflict for long. If you disagree with someone else's results you can go on repeating the experiments until everyone comes to a consensus about which of you was right or until some crucial difference in the experiments is discovered. In parapsychology, of course, unrepeatability makes that consensus impossible. You have to agree to differ."
One requirement of the existence of controlled, repeatable experiments is the accessibility of data. After all, it isn't possible for scientists to repeat an experiment and compare that to earlier experiments unless the original data from those earlier experiments is readily accessible. Unfortunately, such accessibility is often not in existence when it comes psi experiments.
This isn't always the case, but it happens too often that parapsychologists do not make their notes and charts available to non-parapsychologists and independent researchers. This is one of the things which often lies at the heart of charges of unprofessionalism which are sometimes directed at parapsychological studies.
Dr. Charles T. Tart - An Example of What Not to Do
A good example of how parapsychologists have failed to properly appreciate the importance of controlled experiments can be found with Dr. Charles T. Tart, a parapsychologist at the University of California at Davis. Tart had reported successful ESP tests which went far beyond anything thus far found - but there were flaws in his experiment which others pointed out. Once he fixed those flaws and repeated his experiments, he got negative results - but he refused to withdraw his original findings, claiming:
"Because the level of scoring in the first experiment was so high, it would be absurd to argue that the results of the second experiment mean that the results of the first experiment were a mere statistical fluke."
Tart ignores the most basic principles of science - something not uncommon in parapsychology. At no point have parapsychologists produced a psychic whose alleged powers can be demonstrated through a series of controlled experiments performed by independent investigators who employ strict standards. After more than 130 years of research, there is no consistent evidence or parapsychological activity which would survive basic scientific scrutiny.
If parapsychology is a science, why is that?