Murphy's Law of Research tells us that "Enough research will tend to support your theory", but would it suffice? People often accept new information only if it confirms what they believe in. At least it should come in a palatable form. In cognitive science this trait is called "confirmation bias". It is also denoted as the Tolstoy syndrome, after the great Russian writer who had already been wise to it in the 19th century.
The 2-4-6 problem, developed by Peter Cathcart was one of the first systematic attempts to investigate the confirmation bias problem, using a simple device. In a Cathcartian test, the subjects are given three numbers. They are informed that the sequence follows a rule; and they are asked to discover the rule and make their own sequences based on it. The numbers they get have an ascending order with equal intervals, as in 1-3-5. And so they tend to think that is the rule, and they make sequences like 10-20-30, 100-104-108, etc.
Actually the rule is simpler than it looks. It only suggests that any three numbers in ascending order is acceptable, but the subjects usually have a problem in discerning it, and few would try negative numbers! Well, what would happen if new information is passed to two groups of people who differ in their bias. Their views can move even further apart. There is a term for that too: "attitude polarization".
What is your take on it, dear researchers?