Researching police procedure etc. would still be following scientific process, wouldn't it? The very fact that, when writing, you want something like that to be correct, betrays your inner amateur scientist.
A good exercise they do for design courses in art colleges is to take a product or idea right back to the beginning. You do this by removing the established name, so a "kettle" becomes a "pourable container for heating liquid" and a "lightbulb" becomes a "electrical fixture for generating light". Seems like a silly exercise but stops us making assumptions and, in writing, probably allows for a lot more creativity. Why do these aliens who have never been in contact with us have the same teacups, drapes, rugs. Why has their vegetation evolved in such a way that their homeworld looks like Canada? What the hell's going on?
The same can be said of where we get our information - the real world, and observation thereof, or existing texts (which are authored). A good example of authored texts messing with the real world to make a better story is UK police drama The Bill. In this, when a suspect is getting interviewed, their solicitor is present in the room with them. That's not actually how it works here, but it makes for a better scene while also representing the questioning process in a rosier light. So if you were writing a similar scene, which you hadn't researched properly, you'd probably assume that's how it works. Lazy.