People can't do the hypothetical scientific method that everyone learned in school for every study, because it costs a huge amount of time and money. Most indiv

People can't do the hypothetical scientific method that everyone learned in school for every study, because it costs a huge amount of time and money. Most individuals can't follow 10k people for 20 years and see whether obesity goes away when doctors recommend high intensity interval training vs the normal standard of care. Worse, if you take facts from your life you'll inevitable draw a smaller and more biased sample size. Maybe the people you know are already doing HIIT. Maybe you know knowledgeable/ignorant people who are more/less aware of exercise. Maybe by random chance you know three people who actually listened to their doctor, and nobody who didn't listen will admit it. It's probably better to do the Actual Scientific Method and search Google Scholar to see if anyone made a funnel plot to check for publication bias, if the meta-analyses look okay, if there are replications by other people, if there's an ongoing war between two groups of scientists who both have good points and superficially-similar studies in opposite directions, and basically just trusting that if there are ten studies from different institutions they can't all be committing academic fraud.