Yes, there is nothing surprising or unusual about the fact I sometimes find that only about a quarter of the answers to one of my tests (with four possible answers) are correct.
What you fail to grasp is that this is the effect predicted by the null hypothesis. The proper response from science in this case is to reject the operative hypothesis. It did not predict the outcome. The null was not falsified.
One of the important discoveries that I made early in these online telepathy tests is that the answers should always be analysed and interpreted (this has to be done with care and without bias, of course).
No. This is not a "discovery" on your part. This is not something you should be proud of.
Post hoc interpretation of experimental results is one of the most egregious hallmarks of pseudo-science, and you're wearing it like a medal. The goal of science is to control for and therefore eliminate all such irreproducible elements. You have been shown protocols that do this, but since your claimed skill disappears entirely when those protocols are applied, you find reasons to reject them and rely on your misapplied reputation as a scientist to advocate for them.
I should remind you that you already let the cat out of the bag. You already let slip that you define "good" results as those which support your predetermined conclusion, and "bad" results are those that reject it. You have already admitted that you use a biased interpretation to reject disconfirmatory trials and accept only confirmatory trials.
Yes, but you expend effort toward finding new ways of cheating and trying to pass it off as your brilliant scientific discovery. Real science expends efforts toward eliminating the variables you're strenuously trying to put in.
When the results of an extra-sensory test seem too good to be true...
What do you intend "good" and "true" to mean in this statement?
...this may indicate some cheating has taken place.
No. Real scientific protocols make it impossible to cheat, or to make attempts to cheat objectively evident instead of subject to the feelings of the test conductor. You were shown how this could be done, but you are not interested because you specifically intend to cheat and call the cheating a new important discovery you should be praised for.
When the results lead more to a conclusion like "it's a little hard, but it seems to work", then, this may mean some serious and reliable result has been achieved.
No. The reliability of the result derives from the rigor of the protocol, the degree to which it is followed, and the conformance of the collected data to the statistical model. Scientific validity of the experiment is
never predicated on whether the data "seem" to validate the operative hypothesis. That is straight-up pseudo-science.
...one of the participants, Safraniamagik, stood out because she said, in a completely extraordinary way...
No. Since you knew whether the answer given by the participant favored your desired outcome, it doesn't matter what sham rationale you give for accepting or rejecting the data. Her datum stood out because you needed it to be included in your final data set. Your
ad hoc gut feeling about metadata you should have excluded is not the real story.
Post hoc judgment is untrustworthy
per se, regardless of what you want to argue is "extraordinary."
If I decide to accept as credible only respondents who seemed to acknowledge that some extra-sensory perception has taken place, then I get a 100% hit rate for credible answers.
This is simply
post hoc pseudo-science, Michel. You have stated that you think "good" results are only those results that confirm your hypothesis. Therefore you have invented an irreproducible
post hoc system of subjective analysis that is clearly aimed at producing your idea of "good" results. Your subjective notions of "credibility" have no place in this kind of research and are completely controlled for in a properly modeled study.
...I nevertheless concluded the tests had been successes, after distinguishing carefully between the answers which seem serious and those which did not.
No. You didn't like the results you got, so you applied pseudo-scientific
post hoc interpretation to achieve your desired results.
You seem to think you're very clever for "discovering" that if you manipulate data in a
post hoc fashion, you can make the results come out however you want. You seem to be proud of having discovered a well-known shenanigan of pseudo-science.
This is not science. You are not a scientist. You admit that you have no appropriate qualification or experience in human-subject research. Whereas I'm taking a break to write this from writing a methodology for human-subject research that will be subject to approval from several government agencies who regulate this sort of thing. I'm the principal investigator in the study. You don't know what you're talking about.
