I apologize that this post is rambling, and pulls a bunch of different threads into one cloth, but I can't seem to make it any simpler than it currently is.
In a legal court there is a plaintiff and a defendant. In science there is theory and data. In a legal court the judge and jury can rule for either party. In science data trumps theory every time. It is theory's job to explain the data, data needs no defense.
Last year I was trying to explain to a fellow infantry officer who wears a Ranger tab that while the theories of Dave Grosman (former infantry officer who earned the Ranger tab) fit the data he had when he wrote the book, two decades of new data conclusively prove his theory wrong. The other officer wouldn't hear of it at all, and fell back into the logical trap of "appeal to authority" because he didn't know the data. In his mind the theory was true, and the data was invalid. I know that "perception is reality" but I have to say that people who cling to a theory that is shown to be wrong by the data can only be suffering from "Joan Peterson Syndrome" and not in the amusing Adam Savage "I reject your reality and substitute my own" way.
This gets us to "global warming" where the data does not fit the hypothesis that Carbon Dioxide absorbs infrared radiation and therefore heats the atmosphere. The data that we have does not support the theory (at best the data can be massaged either way). The best data set for this would be weather balloon data because it measures atmospheric heat without confounding factors such as "urban heat island" affects or having an air conditioning unit blow directly on the thermometer as happens with ground based measuring stations. The balloon data shows no warming despite a measurable increase in atmospheric CO2. http://wattsupwiththat.com/2010/03/18/weather-balloon-data-backs-up-missing-decline-found-in-old-magazine/
This also gets us to "Creation Science" where the assumption is that the writings of bronze and iron age feudal tribes are true, and the goal of "research" is to support that truth. Obviously that isn't science, and while it is interesting (if you want to find all the holes in Darwinism, Neo Darwinism, and whatnot the Creationists are an easy place to start).
Global warming alarmists have taken the same stance, that an unproven theory proposed by a man regarded as "a lunatic" by his peers in meteorological research in the 1970s that Carbon Dioxide might be able to stop the onset of global cooling is true.
Tam and others have noted that people get their ego, their self identity, built into any new gun they purchase. If you hang around the internets long enough you'll see plenty of "I just bought the Bangwiz 30k, tell me how awesome it is" for those who are seeking validation of their choices, which means validation of their entire being.
If "global warming" were truly science then scientists would scrap the theory that CO2 causes atmospheric temperature increases due to absorption of radiation in the IR spectrum and propose a different hypothesis. Oftentimes a "null hypothesis" is part of a well designed experiment simply to give a balance in statistical data. If you look at the data, which is more true, the hypothesis or the null hypothesis (and I apologize to all mathematicians for the gross simplification).
However, there are those who suffer from "Joan Peterson Syndrome" and refuse to accept that the null hypothesis is valid and the hypothesis is invalid. Now we could say that this is a simple mental problem, if it were not also accompanied by an unhealthy sociopathic desire to act on this false reality and impose their will on other people. And this is where the rubber meets the road, in control. Joan Peterson wants to take away your guns and create a police state, Al Gore wants to make you live like a third world peasant in a police state, and Dave Grossman wants your kids to not have graphic video games (Dave may be correct on other aspects of his work, but he is clearly wrong on violent video games turning kids into highly trained violent commandos).
But back to science and the courtroom. Who will win in theory v. data is largely up to the side that invests the most. After all, how many millions have died trying to make the lies of Communist theory true? The data says otherwise, but people suffering from "JPS" are more than willing to give it one more college try, no matter what the previous data says. Seriously, do you want crazy people to make decisions that affect everyone based on their inability to determine truth from falsehood because it would destroy their very identity as a human being?
What is worse, is that people looking for data to support their "theory" that they have already decided is true is rampant even among those who should know better. This is why scientists discuss whether an experiment is "well designed" or not. A well designed experiment avoids the pitfalls of small sample sizes, unambiguous data sets, and a control experiment. A "well designed experiment" is above all else one that can produce repeatable results. The "Peer Review" process is supposed to weed out poorly designed publications, unfortunately that doesn't always happen.
When the theory can't produce repeatable results (or produces consistent results that disprove the theory) then you need to scrap the theory, not cling to it bitterly. But from "gun control" to "global warming" to "violent video games" to "socialist paradise" we see that people are incapable of abandoning a known falsehood.