This is a hard problem. At some level, all people have to rely on experts to interpret scientific results for them. In science, the standard of truth is experimental repeatability, and most results that are interesting enough to get published require quite a lot of effort to actually run the experiments. And at some level you either have to run the experiment for yourself or trust somebody's word that they got the result they got and didn't make any mistakes.

So, since replicating results is hard we rely on reputation and authority instead. Is this paper peer reviewed and accepted to a respected journal or conference? Are respectable journalists writing about it? Are government officials saying things that match what the paper seems to be saying? And so on.

There are shortcuts. If an article is claiming that scientists have created a perpetual motion machine, they haven't. If an article was plainly written by someone who doesn't understand the difference between a kilowatt and a kilowatt-hour, it can safely be ignored. And so on. Science literacy is helpful at spotting obviously wrong articles, but I think there's a danger that it can lull one into overconfidence in one's own abilities to spot falsehoods. Sometimes the things we're told are just plain wrong, and some of those things aren't obviously nonsense. Some of them come from seemingly reputable sources and could plausibly be true, and the only way to know for sure is to see if other reputable scientists are getting the same results.