Very easy to do by default if you haven't done enough historical reading -- especially of primary source -- when you don't realize that what comes natural to you is not the nature of the universe. And a grave danger when dealing with eras where we have very little documentation so it's hard to piece together. However. . . .
Sometimes I suspect that some writers pick eras that are very little documented to put their stories in (or scrub off a few serial numbers and then put their stories in), because they can paint their own notions Very Large over the wide expanses of blankness. It can be very sad when a writer doesn't realize that an era is more documented than he realizes -- Ancient Greece, for instance -- and starts painting notions over things we know. It can be very annoying when a writer takes advantage of the illiteracy of barbarian tribes to imagine they are different from any other barbarian tribe has ever been and instead like twenty-first century urban America.
Of course, this may not be the writer I read that's at fault. Secondary sources, which may be all that the writer had to go on, often plaster their own notions very large over the material, and it takes a keen eye to ferret out modern day from actual evidence. (Good secondary sources discuss how they make deductions from the evidence. This is useful training for this purpose.) Then it lasts forever. I have read writers who parrot the wild guesses of Victorian anthropologists -- often extremely bigoted ones, whose views were as much devised to hurt as to be true -- with a childlike trust in their accuracy.
This -- The Myths of Avalon -- is a good article on how Celtic history and culture has been mythologized. And the inspiration for this mediation on history and obscurity. 0:)