PZ Myers had two such cases he posted about today. The first being a case where a woman studying rape cases concluded that men that chased a lot of women where more likely to commit rape, which got "turned into" an article in the British Telegraph, stating that the study implied that women that dress provocatively where more likely to be victims. The second was a case where the "editor" of a journal cut an 800 word article to 360 words, and in the process "changed" key details about the lack of similarity between human ears and fish "lateral lines", which could destroy the young journalist's reputation with the person he interviewed, who specifically told him they where *not* similar. Another person, who works as a science journalist, relates a tale in the comments about how something similar took place, and the person he interviewed nearly had him black listed, until he cleared up the issue. I.e., the editor nearly destroyed his career by making the changes, then publishing, without bothering to check that they hadn't changed critical details. Who knows how many people may now be flipping hamburgers, who went into journalism, because the editors and other people that run papers and journals have less competence than even the journalists, and ended someone's career this way.
And, that isn't even mentioning papers, journals, and news agencies, whose reporters and editors are *specifically paid* to distort things to fit their ideological positions.
Scientists do have some issues with communicating clearly, but when they get interviewed by people that already know what they "want" the story to say, or don't have, nor want to take, the time to understand what is being said, you get bad journalism. And, if you then happen to be working for a paper/journal/magazine that won't double check with the journalist, never mind the scientist, before going to print, to make sure the article is accurate... How the heck is the public supposed to have a clue what is really going on?
Scientists are doing their job, which is doing science, not writing journal articles, or making sure that the people that do don't screw up and make false claims about what they are doing. You can't expect them to stop doing a 6 month study, to spend a week fixing what some halfwit mangled in an article. What you can expect, and the public should demand, is that journalists, editors and the people running their news sources, report the most accurate information they have, and double check it first, even if that means missing some stupid deadline, instead of worrying about getting it in print immediately, or looking for some grand, entertaining, and completely ass backwards, narrative. We are not publishing fracking Science Fiction here, we are publishing Science. And the later requires accuracy ***first***, not as an after thought.