Null probability of change among zealots


Learning the language of science is challenging. As the first hurdle, one has to accept that science seldom reveals any deep Truths. Rather, scientific findings are almost always couched in probabilities. These statements of probability usually go something like, “It is more probable than not that Event X influenced the observed change in Condition Y.”

Few of us rush out with laurels and trumpets over statements of this kind.

What makes science even less digestible is a thing called the Null Hypothesis. The Null Hypothesis is a kind of theoretical construct against which scientists test their ideas as to how things work. In many cases the Null Hypothesis is a statement like, “The observed change in Condition Y is due to measurement error or random chance.” In other words, scientists ask: How likely is it that the thing I think caused the change actually caused the change… or is it more likely that either random chance or errors in measurement caused the observed outcome?

Hence, probability statements. If the measurement is good enough (typically a 95 percent chance that the thing did what the scientist thought and that it was measured properly) then, one can “refute the Null Hypothesis.”

While space will not permit even a brief description of the intricacies involved in the math of probability calculations, suffice it to say they are complicated.

Even if a scientist refutes the Null Hypothesis, he or she hasn’t necessarily proven anything… aside from the likelihood that their idea may have some merit. Again, few laurels and trumpets…

Nonetheless, this is important. We all make important daily decisions based on probability structures. We just don’t think about it that way. For instance, you don’t run in front of speeding buses because you have calculated that the probability that you would be grievously injured is greater than that of simply being gently swept aside.

This is the fundamental nature of empirical observation. We gather evidence —- I saw the fool on the evening news who tried to beat the bus. I reasoned, based on that observation, that certain courses of action, namely bus charging, are probably not in my best interest.

This is what scientists do. The only difference is that their gathering of evidence is far more systematic and far more explicit. There is very little black box magic in mainstream science. Just as in eighth grade math, they have to show their work.

Not only do they have to show their work, but if they want it published in credible outlets, their work goes through a process called blind peer review. In this process, other experts in their field (neither party knows who the other is) review the submitted manuscript and respond with comments and critique. It is a humbling and often frustrating process. Typically there are at least three reviewers and you have to get at least two of them to believe in the strength of your work.

Does this mean that every paper that gets published is good science? No, far from it, but therein lies the elegance of the process. Once published, other scientists try to replicate the first author’s results. If they can’t, then often a second “response” paper (or a bunch of them) gets published.

All of this gets to a simple point: when thousands of scientists agree that something is more likely than not, it actually means that tens of thousands of scientists have reviewed the work and assented to its merit.

Recently, James Powell, published what’s called a “meta-study” of articles examining climate change. He looked at all the peer-reviewed scientific articles on climate change between January 1, 1991 and November 2012. Of these 13,974 articles (written by a total of 33,690 individual authors), exactly 24 reject global warming or endorse a cause other than CO2 emissions for observed warming. That’s a ratio of 581:1. Nobody would bet on that horse at the track. Even so, there are a lot of folks out there who try to portray a divide among scientists on the topic.

There’s a divide all right —- about as wide as a gnat’s crack. There are also Holocaust deniers and people who dispute evolution. Unfortunately, we manage to regularly elect them to Congress. Believe whatever fantasy suits your political agenda, but be big enough to call it what it is… ‘cause it sure ain’t science.