Feedback Biases

Science: Experiments often give contradictory results. Whether measuring the charge of an electron, the age of a piece of rock, or the effects of the South Beach Diet on health, every run of a test will give a slightly different answer. This is to be expected. Where the problem lies is in what the experimenters do with the results. Often, an experimenter has some prior expectation of about what the result should be (especially if the run in question is not the first). When the result comes up outside the expected parameters, the experimenter will check the apparatus carefully to make sure it is functioning, the samples to make sure they are pure, and the math to make sure it was done correctly. If a problem is found, the result is ignored and not published (makes sense so far). However, when a result comes up within the parameters they expect, nothing is checked for problems. Why would there be a problem? Everything came out as expected. When things go wrong, they look to see what went wrong, but when things go right, they don’t look to see what went right. This is an example of what is called a feedback or confirmation bias.

​One may wonder how we can trust our scientific establishment because of this. Actually, the science community defends these actions. I once heard a science teacher state that because of peer-review processes and independent experiments by others, this form of bias is kept in check. After all, that is how science has been done for a long time – but this is no solution. The other experimenters are also biased, often towards the same objectives. Simply being scientists gives them too much in common to be truly independent. Living in the same society, or even on the same planet, or even belonging to the same species means that many ideas are potentially embraced by a majority of experimenters. Many ideas are very common, such as inflation theory, superstring theory, neo-Darwinian evolution, punctuated equilibrium, intelligent design, Freudianism, democracy, Islam, and communism. Those on the fringe that do not share these beliefs are likely to be ignored simply because they are in the minority. In this way, dominant theories dominate more and more in a feedback loop. In the world of ideas, the rich often get richer while the poor get poorer.

The News: The same type of problem exists outside of formal science as well. How we conduct our lives is based on the same circular reasoning. We believe rumors if they seem to be consistent with what we already believe, which is also a bunch of rumors. Most of you have probably heard that we have a problem with frivolous lawsuits in the United States. Some of you may actually remember individual cases. A commonly known one is the suit against McDonald’s because some lady accidentally burned herself with McDonald’s coffee. Common people were upset that any judge would judge in favor of the plaintiff in such a case. McDonald’s was not to blame for her clumsiness. They were not surprised, however. Many had come to believe that the legal system was becoming more and more inane. Because they were already predisposed to believe that these types of cases were common, the story of this woman and McDonald’s was not doubted, and the facts were not checked. The story spread quickly and became commonly known.

As it turns out, there is some dispute over what actually happened. According to one source, McDonald’s was only judged half responsible, and that was only because their coffee was kept at much higher temperatures than industry standards. It wasn’t just hot coffee; it was dangerously hot coffee. The point is that it is possible that many (or most) of the stories that we read about in the news and hear from friends are incomplete and misleading. The reason they are accepted as truth is because they are believed to be common, yet the reason they are believed to be common is because there are so many stories. The reason this is a problem is that some politicians have suggested reforming the judicial branch to deal with these lawsuits. Whether or not there really is a problem becomes an important factor in determining what actions to take to fix it, and fixing a problem that isn’t there may instead cause a bigger problem.

A related disinformation problem is determining how common is too much. There will always be some abuses of our legal systems, but at what point can it be said that we have “a problem?” People assume that the stories they read about are just a small sample of what is out there, meaning the totality is much larger, but suppose that what we see is the totality of it. News agencies want to print interesting news to keep their customers. Juicy stories of frivolous lawsuits are a favorite. It could be that there is no major problem, but only that every single abuse that ever happens gets to every listener from reporters in the media and then also by word-of-mouth. This does not mean that all our common beliefs that we receive as rumor are false, but many of them might be.

Feedback Spirals: If you are scientifically curious you may have wondered how scientists know how old various layers of rock are. They do some radiometric testing, but this is costly and difficult. Mostly, they check for fossils. If they see a fossil known to be from a certain era, they then know that the rock surrounding it is also that old. And how do they know what era a fossil is from? Usually they check the recorded age of the rock layer it is in.

Wait a second, isn’t that circular reasoning? That’s what young-earth creationists claim. In fact, this is one of their main talking points they use to convince people that evolution theory is in error, and the evolutionists do not deny that they engage in this practice! Rather, they claim that instead of a feedback loop, it is more like a feedback spiral. Those fossils first discovered had their surrounding rocks tested radiometrically. The estimated rate of sedimentation and other clues corroborated these dates. It was only then, that rather than go through the lengthy and expensive procedure of radiometric testing, scientists began to determine the ages of fossils by the known ages of the rock they were found in, according to earlier tests. This repeated activity built up a database of known fossil ages, which were in turn used to determine rock strata ages that had yet to be tested. To further clarify: suppose that it was found early on that trilobite fossils were only found in rock strata from a specific, narrow span of time. Then, anytime someone found a new layer of rock with a trilobite fossil in it, they would assume that the rock was also from this time span.

The young-earth creationists, of course, point out that the number of trilobite fossils collected during the early days of paleontology may have been too small, statistically speaking, to rule out their prevalence elsewhere. If a large percentage of trilobite fossils exist outside the range assigned to them, the rock strata dates derived from them would then be horribly flawed. The flawed rock dates would then corrupt the fossil dates derived from them, which would in turn corrupt the rock dates derived from those fossils, and so on. In fact, young-earth creationists even point out that some trilobite fossils have been found over (and therefore in younger strata than) dinosaur fossils (which were supposed to have evolved later). These rare events are easily explained away by evolutionists as folding and overturning of the crust, which is agreed by all to happen sometimes. The disagreement is over whether it happened in those specific instances.

It wouldn’t even take very many out-of-place trilobite fossils to screw up the whole system. If just one is slightly above its prescribed zone, it would give that rock layer a slightly older date, in turn giving all fossils in that layer a slightly older date. If every group of fossils overlaps slightly with every other group, it could be theoretically possible to assign all fossils to the age of trilobites. Of course, with various cross-checks this is highly unlikely in a complete and rich enough fossil record. However, nowhere on Earth can the entire fossil record column be found. It has been pieced together from different places. Add into this mix the tendency to doubt one’s individual dating results and look for any explanation to discredit them rather than doubt the whole theory, and a potent combination has been made in which confirmation bias can flourish. It would seem that “feedback spirals” can very easily fall into the trap of circular reasoning, making them feedback loops after all.

Wisdom: Nevertheless, despite this tendency we cannot assume that this feedback spiral has degenerated into a loop, or that one hundred and fifty-eight years of well-established evolutionary theory are wrong. When they have a proper foundation, feedback spirals are legitimate constructs. I don’t know what to think. I wasn’t around during creation to see how God did it or how long he took. Is the Earth 4.5 billion years old? Quite possibly. I just want to open minds to understand how unsure the intellectual elites really are about everything. The same principles apply to nutrition, the environment, and economics. The wisest man is the one who knows that he knows nothing.

What do you know?