How can we ever be sure?

[T'ai Chi]

In regard to anomalous cognition experiments, in one of Hyman's writings he asks: "Has the possibility of methodological flaws been completely eliminated?"

My question is, is can the possibility of methodological flaws ever be completely eliminated in any experiment?, anomalous cognition related or otherwise.

[neutrino_cannon]

I'm sure you had the answer to the question already, no we cannot be sure. Of anything. Ever.

Which is no reason not to experiment. Inquiry to the point of reasonable assumptions seems to work pretty well.

[Darat]

Originally posted by T'ai Chi
In regard to anomalous cognition experiments, in one of Hyman's writings he asks: "Has the possibility of methodological flaws been completely eliminated?"

My question is, is can the possibility of methodological flaws ever be completely eliminated in any experiment?, anomalous cognition related or otherwise.

Well peer-review helps by having several people check the work etc. (But peer-review is quite often limited in not being able to check the actual experimental setup rather then the reported setup.)

But the process of science with its "principles" of repeatability, revealing all data and so on helps to ensure that even if methodological errors are present in one particular experiment they are "teased" out in future experiments. And this is why just one experiment is quite often not enough.

It is hard to comment further on Hyman's particular quote unless you can provide the context.

[Soapy Sam]

Originally posted by neutrino_cannon
I'm sure you had the answer to the question already, no we cannot be sure. Of anything. Ever.




Uh...how can you...? no, no. Forget I asked.

[Paul C. Anagnostopoulos]

It would help quite a bit if the hypotheses in psi experiments were derived from some theory of psi. Instead, they are about the statistical outcome of the experiment itself. They are anomaly searching hypotheses.

If the hypotheses were about a theory, then people could develop additional hypotheses from the theory, and they could develop alternate protocols to test the same hypotheses. This would have the benefit of changing the potential methodological flaws. If the results were still positive, that would indicate that the results were not due to flaws. Instead, changing protocols tends to eliminate the result, as was so clearly shown in the autoganzfeld experiments.

I think there is a trend toward trying to come up with theories and testing them. That is a good thing, as long as experimenters can find some psi effect that is reproducible enough to test from many different angles.

~~ Paul

[epepke]

Originally posted by T'ai Chi
In regard to anomalous cognition experiments, in one of Hyman's writings he asks: "Has the possibility of methodological flaws been completely eliminated?"

My question is, is can the possibility of methodological flaws ever be completely eliminated in any experiment?, anomalous cognition related or otherwise.

Of course they can't.

However, many experiments show an effect that you can see with your eyes closed.

These experiments of which you speak tend only to show results that are just above the threshold of statistical significance, and that only if you squint your eyes and hold your head funny.

It is perfectly reasonable to look closer for methodological flaws in experiments that produce, at best, marginal results versus ones that are like an ice pick in the forehead.