Something people don't consider is that studies often are cherry picked. Meaning if they don't got the result they want, they will discard the study and you will never hear about it. Drug studies can be fudged this way. They can do 100 studies, disregard 99, and report on the 1 that seemed most promising.

Academic fudging is real. I think something like half academic papers can't be reproduced. Imagine the low standards for this kind of research. And even lower standards for news articles.