The title: IF YOU BELIEVE IN CONSPIRACY THEORIES, SCIENCE SAYS THERE’S SOMETHING WRONG WITH YOUR BRAIN
The typical way to lie with statistics.
Let's look, first, at the facts presented in the article. There is some "mental disturbance called “illusory pattern perception.” This is defined as the mind’s ability to identify “a coherent and meaningful interrelationship among a set of random or unrelated stimuli.” In other words, finding patterns where none exist."
First of all, how problematic is this? Problematic enough to be named "mental disturbance"? The ability to identify patterns is an extremely useful thing, and almost a necessity for a successful scientist. And it does not really matter if there is some overproduction of such pattern identification. What distinguishes the good scientist is not that he has no such overproduction of patterns, and does not see patterns if there are none in reality, but that he is able to evaluate the patterns he observes and finally recognizes and rejects the wrong ones.
Let's take into account also the dangers of error: If you identify a wrong pattern in a random sequence, the harm is not big. If you don't recognize a meaningful pattern, the harm can be quite large, if not deadly. So, from point of view of survival, it makes a lot of sense to err on the side of overidentification of patterns.
But, of course, everything can become extreme, and in some extreme cases too much of such misguided pattern identification may be worth to be named a "mental disturbance". But read the description of the experiment: Nothing suggests that what was measured was such an extreme overproduction of pattern identification. One can, of course, say that a pattern recognized in a random sequence is wrong, but so what? This is the usual way pattern recognition works: We start with identifying patterns. After this, we start to test if there is really something behind this, by testing if the same pattern is repeated in our future observations. Usually it is not, and we throw away that pattern as a funny accident, and forget about it. This normal way to identify meaningful patterns essentially depends on the ability to overidentify patterns, to observe some even in random sequences.
So, this is manipulation 1: Name something quite normal and unproblematic a "mental disturbance".
Of course, it is quite plausible that this particular "mental disturbance" is correlated with belief into various nonsense theories. Ok.
Manipulation 2: The belief in such nonsense theories is combined with belief into theories which are politically unwanted, by naming both "conspiracy theories". Some statistical connection will remain between those witha "mental disturbance" and those believing in what one has named "conspiracy theory". Moreover, it is quite plausible that good pattern recognition (the "disturbance") also makes it easier to identify government lies.
Manipulation 3: Transform a statistical correlation (however weak, if you can name it "statistically significant", it sounds as important) into a causal, one-sided relationship:
If you believe in conspiracy theories,
then you have a "mental disturbance".