JohnStOnge, you only posted other people's opinions about Popper, you didn't post anything from the man himself. Of course some today are trying to discredit him, because they are precisely the type of people he was trying to put in their place in his day.
Ok, I went ahead and read an essay on the subject by Popper himself. It's at
http://stripe.colorado.edu/~fredrice/popper.pdf . In my opinion his philosophy was pretty much accurately represented by the sources I'd looked at.
If all he said is that a theory has to be testable, I'd agree with that. The scientific method clearly requires that a hypothesis has to be tested through experimentation before it can become a theory. But that's not all he said. He wrote this:
"Every genuine test of a theory is an attempt to falsify it, or refute it.
That's simply not true. Take the hypothesis that putting mice on a low calorie diet results, on average, in greater longevity. This is how experiments to test that theory will generally proceed:
Mice will be randomly assigned the treatment of reduced caloric intake. Control mice will be randomly assigned the status of being controls. Longevity will be measured. A statstical test will be applied to the results. If the result is that the mean longevity of the low caloric intake mice is "significantly" greater at an acceptable level of confidence (usually 95%), the outcome will be considered to support the hypothesis. After that's done a bunch of times it'll be considered a very strongly substantiated theory.
There is NO effort to refute the hypothesis during that process. In fact, if the truth is that a low calorie diet has no effect, the hypothesis CAN NOT be refuted because the null hypothesis CAN NOT be inferred as true. Though you see it a lot, it is improper to say that ANY statistical experiment showed that there was "no difference" or "no effect."
What's going on is that the investigator sets up a conceptual model in which the treatment has no effect. That is the "null" hypothesis. The alternative hypothesis is that the treatment does have the hypothesized effect. The alternative hypothesis is inferred if the results are such that they would not be likely to have occured if the null hypothesis were true. There can be two outcomes to the experiment. The one the investigator is hoping for is stated something like, "There is sufficient evidence, at the 95% confidence level, to conclude that the altenrative hypothesis (low calorie diet means greater longevity on average) is true." The other outcome is "There is NOT sufficient evidence to conclude, at the 95% confidence level, that the alternative hypothesis is true." But that does not represent having refuted the alternative hypotheis. If the null hypothesis is "no effect,"
it is not possible to infer it as true. It is only possible to infer it as false.
The only way the alternative hypothesis can be inferred as false in such an experiment is if the results come out so that it looks like the effect is actually opposite from that hypothesized. So if you wanted to you could set up an experiment and make the alternative hypothesis the opposite of what you actually want to infer. You could say your null hypothesis is that there is either no effect or that mice live longer on low calorie diets. Then the alternative hypothesis would be that mice don't live as long on low calorie diets. And if you got results unlikely to have occurred under the null hypothesis you would have falsified the hypothesis that mice live longer on low calorie diets.
If you accept Popper's idea that any good test of a theory is an effort to refute it, that's what you'd have to do. But, if your hypothesis is that mice live longer on low calorie diets, that makes absolutely no sense. The idea is that you have to "prove" that they live longer on low calorie diets and the way to do that is to set your null hypothesis up as "no effect or they don't live as long." And when you do that you are NOT trying to set your experiment up to refute your hypothesis.