Originally Posted by genejockey

No, but that's how science works. If you wore a base layer on a hot day and felt hotter and sweated more than you did on a hot day with only a jersey, that would disprove the hypothesis, wouldn't it?


Nonsense! I proposed the same rider try riding with and without. That eliminates the variable of the rider right off the bat. There's still the subjectivity of the observation, but you could go out on days that are the same temperature, riding the same route at the same speed - or ideally the same power output at all points, and observe how much you drank and your weight before and after each. Do this repeatedly and you build up significance, just as you would if you had more riders conduct the same experiment.



True, but disproving it is as far as you need to go in proving it to be wrong.



Sorry to disappoint you, but that is exactly how all good scientists work.

You come up with your hypothesis, and then you try as hard as you can to prove yourself wrong - "If this hypothesis is true, then If I do X, the result should be Y." If the result is Y, you still haven't proved your hypothesis, you simply failed to DISprove it. It doesn't mean it's true, it just means you've now know that it's LIKELY to be true. You keep on trying to disprove it as better tests become available, and maybe you finally reach the point where it fails a test, and then you refine it, and test the refinement. Part of this is the knowledge that other people will also be trying to disprove your hypothesis, and it's better for your ego and your career if YOU do it, rather than somebody else.

This is the entire point of p-values - they tell you how likely your result is to just be by chance. We generally accept results that have less than a 5% chance of being wrong, but that means we still can be wrong 1 out of 20 times. That's science. The lower the p-value, the more likely you're right, but it's never zero.

The courts work on the basis of who makes the most compelling argument to either a judge or a jury. If that proved things, there'd never be reversed verdicts, and nobody ever exonerated.