Thanks Cassie for a useful article. I've found it's worth bearing a few things in mind. First of all, science is never about belief. Science is about trying to derive adequate explanations of repeatable phenomenon that have predictive power. Through science we learn our way forward into the complexities of the real world, and we revise our ideas as we discover more phenomenon and develop more adequate theories. Secondly, all science is done by people and therefore is fallible. Both Science and Nature back in 2006 carried articles about how at least 50% of all research published in highly reputable journals is in fact total junk. There are, as you explore, many reasons for this. Therefore we need to see repeatability by independent teams before we take a result as being an indicator of something.
Finally, we must always remember that mass media headlines are sensationalist and rarely if ever convey what the study has actually found. Always read the original paper, never the vapid summary by a know-nothing journalist desperate to grab eyeballs.