We all have pre-conceived ideas, and convincing evidence in support of those ideas is readily accepted. Evidence to the contrary is easily minimised, ignored or dismissed.
For instance this data from India is compelling, but it was never taken seriously. Note the date, 1960's. The heart disease rate was seven times higher, in the vegetarian, low-fat, south of India. There is a cultural barrier, it's about India, Americans are saying "it's not about us," as if the people of India, and science from India, are not to be trusted. There's a cost to that.
Sometimes people prepare elaborate defence's of their existing position. The Stellenbosch University - Cochrane Collaboration usually called the Naudé Review is a case in point. They compared several low-carbohydrate diets with a standard diet and claimed that for weight loss there was no difference. The study is significant for what it didn't do. The study defined low-carbohydrate as less 40% of total calories, about 220gm of carbohydrate per day. They did not do a comparison with very low-carbohydrate diets, less that 50gm of carbohydrate per day. Their study by design, didn't look at very low-carbohydrate diets like the Banting Diet. (Probably because the Stellenbosch researchers, and many researchers around the world, are lipophobic, and cannot recommend a high fat diet to trial participants.)
On publication, in July, 2014, the lead author of the Naudé Review, Celeste E Naudé, was active in the South African news media, using the study as evidence to discredit Dr Tim Noakes, The Real Meal Revolution and the Banting diet. Given the limitations of the study this attack was unprofessional and dishonest.
We can all be guilty of allowing our prior knowledge to blind us to new thinking. Commitment to well established conventions, professional standards if you like, can cause that sort of blindness. Loyalty to a convention can trap us in the tradition of the past, and cause us to ignore highly relevant data that should compel us to action.
That applies to each of us as individuals. We can be told what a better diet is, but we won't believe it, if we already "know" what a better diet looks like. That also applies to institutions, where the standards are written down and agreed to, and supported by members. When it becomes clear that something is wrong with those standards, making a proposal for change is both personally difficult, and politically dangerous. All our institutions and political parties suffer in that way, the rate of real change is quickening. The slow response of our institutions puts them under constant pressure.
John Stephen Veitch