"I thought experimentation was a phase that followed observation, in which the experimenter manipulated the conditions to test for the effects of a variable. Maybe this view of experimentation is now outmoded, but I think not."
No, it's not outmoded; I do it all the time. But pure observation is also legitimate evidence, whether it can be followed up by experimentation or not.
Just so I do not misunderstand, are you suggesting that unless data is collected by experimentation it is not legitimate data?
Kevin L. O'Brien
PS: How does bacterial antibiotic resistance falsify the prediction of evolution?