Jones Beene wrote: > > Much of this goes back to the "expectancy effect", expectation-bias or > Tiller effect, which we have all commented on in the past. This is related > to other delusions that afflict even the smartest of us: the Plecebo effect, > the self-fulfilling prophecy effect, or the Pygmalion effect: all of which > are deeply ingrained into human nature. >
Of course the same thing applies from the "other direction", too. At this point, after all those null results, Scott must be *expecting* to get a null result in each new experiment -- at least, if he's any sort of normal human. Consequently he's more likely to be suspicious of the calorimetry, and work hard to fix it, if he's seeing an OU result than if he's seeing a null result! And that'll tend to skew his results in the "null" direction. This brings up an interesting question: Suppose for a moment that the CF results were all errors. Then, that would make me wonder -- is there some global, overarching reason why erroneous calorimetry would tend to OVERread the energy produced? And if not, if erroneous calorimetry results should be randomly distributed, *where* are all the under-unity results? With all those "bogus" results, really, half of them should have shown heat consumption, rather than heat generation! Are heat-deficit results just thrown away out of hand, as being "obviously wrong"? Or does this suggest that the extreme excess of excess heat results over heat deficits must mean there's really something there, after all?