Yes, and my example proves the point I am making -- one must have tentative, falsifiable explanations in science. Choose your own example if you don't like mine, but make no mistake about it -- scientists explain things. If the explanations fail, then another explanation is tried. But description is not science, that's stamp collecting.
This goes as far back as Francis Bacon, who first articulated it -- reading Bacon's philosophical works, the role of explanation, of theory, is clearly set out as a requirement.
More recently, falsifiability has been made a requirement for science, especially in cases where science is defined in legal actions. And descriptions cannot be falsified -- if I say that I saw many tiny points of light in the night sky, that can hardly be falsified. But if I make the claim that those points of light are actually thermonuclear furnaces at a great distance, that explanation is open to examination and falsification, and I have crossed the threshold of science.
This is why the dried gourd example is perfect for my purpose -- it aptly caricatures real psychological work. And once one tries to take a step toward science, to explain the result, it falls apart, like so many psychological studies do.
You mean by quoting authoritative sources like the chairman of the NIMH, whose views resemble my own?
> ... but I think I've got to give up on your posts.
Feel free, but if you've managed to miss the historical change taking place in mental health right now, then nothing I might say will help. The tl;dr is that neuroscience is taking over.
You explicitly stated that the example was "for those who doubt the central role of explanations in science". That you can throw together an example with a bunch of flaws that produces a flawed result is both unsurprising and not especially damning of any specific one of the flaws - one flaw is enough to be wrong. Therefore, your example is not very persuasive on the issue you claimed it directed at.
If you're just trying to get your jollies lampooning everything that sometimes goes wrong with psych research, that's fine - but present it that way. If you're trying to help people reason better - and possibly find flaws in your own reasoning - isolating particular problems is a much better approach.