There is an interesting discussion here which covers the meta-analyses of the ganzfeld studies. It has been determined that the ganzfeld results, even when controlling for or eliminating studies with suspect methodology, are quite satisfactory from a statistical perspective.
It is also the case that tests run in the 1990's and meta-analysis on those tests resulted in poorer results. There is a discussion of that here. At first glance this might be assumed to be caused by the application of superior and more rigorous methodology, and thus showing that earlier researchers were either biased or did poor science. That turns out not to be the case. Upon analysis of the root causes of the change in results, it was determined that the same kinds of experiments were not being performed. A relevant quotation from the 2nd link is: "Many psi researchers believe that the reliability of the basic procedure is sufficiently well established to warrant using it as a tool for the further exploration of psi. Thus, rather than continuing to conduct exact replications, they have been modifying the procedure and extending it into unknown territory. Not unexpectedly, such deviations from exact replication are at increased risk for failure. For example, rather than using visual stimuli, Willin (1996a, 1996b) modified the ganzfeld procedure to test whether senders could communicate musical targets to receivers. They could not. When such studies are thrown into an undifferentiated meta-analysis, the overall effect size is thereby reduced and, perversely, the ganzfeld procedure becomes a victim of its own success."
The whole field of ganzfeld research is very interesting partially because it become a litmus test of scientific inquiry itself. I am fully in favor of strong and stringent standards, but one thing that has become glaringly evident is that when the field of psi research is compared to the fields of research of other "soft" sciences, it appears that the really stringent conditions apply ONLY to psi research. The rest of psychology, and sociology get a complete pass -- those studies are NOT subject to the same rigorous analysis. As an example of the level of rigor that is applied to psi research, see this quote: "Have the ubiquitous doubters been swayed by Honorton's experiments? Some critics of parapsychology, such as S. Blackmore, opine that Honorton has come up with best best evidence yet for telepathy; but Blackmore still has her doubts. Already experimental flaws have been pointed out in Honorton's work. For example, the researchers scoring the experiments must be completely ignorant of which film clips were used, but surreptitious peeks at the automated equipment were possible, and there could have been subliminal cues as to film-clip identities from the time periods required to rewind the tapes. Then, in the scoring conferences with the receivers, the scorers could have subconsciously led the receivers along."
Now, think about this. It is possible that the researchers could have surreptitiously peeked, and there could have been subliminal cues with which the scorers could have subconsciously influenced the experimental subjects. Yes, true -- but is that same level of rigor applied to other experiments in the soft sciences? NO!!!
From the whole area of ganzfeld research I draw 2 conclusions -- 1st, that the level of rigor in the experiments in the soft sciences needs to be increased to the level that the ganzfeld experiments are expected to follow, and 2nd, that to a lot of ganzfeld critics are like kids covering their ears, shutting their eyes, drumming their feet on the ground and saying "Nyah nyah nyah I can't hear you!".
Oh, don't get me started. Forget experimental psychology; I'm highly critical of the lapses of experimental rigor in medical research, ostensibly a much harder science. I'm forever seeing the wildest conclusions confidently based upon the shakiest statistics. (Witness the power line/cancer farce, for example.)
I've only had a few minutes to read up on the "ganzfeld" experiments, but one thing struck me as odd. They have a judge look at four video clips--one of which is the target clip--and compare his impressions to the "receiver's" impressions. What perplexes me is that they refer to the three decoy clips as "controls". They are not controls; they are part of the same experiment. The probability of one of them being selected is not independent of the probability that the target gets selected. A true control would be to preselect one of the clips as the "target" clip and (unbeknownst to the other participants) NOT show anything to the "sender". I don't see where this has been performed.
As Prof. Max Gottlieb always said: "Vere iss your control? Vere iss your control?"
[Hypothesis: the "psychic"--actually, empathetic--link exists not between the "sender" and the "receiver", but between the man who selects the video clips, and the judge. Suggested control experiment: for a random sample of trials, switch the "receiver" impressions around between the trials, and see whether the effect goes away. Never reuse video clips, in any case.]