More than that, even if you are not trying to do bad science, there is always incentive to report bad science. I saw a great example myself a while back, big headline about marijuana use increasing the risk of heart attack.....as a pot smoker I was concerned, so I dig in....
First, it wasn't the main point of the study. Pot smokers made up a small portion of their study population. The overall study was good.
So out of a study of many hundreds, the big headline was on a small sub-population of somewhere around 20 people. The main metric they used was how many hours it had been since a person last smoked.... 24 hours being the lowest.
So basically.... a small number of pot smokers who had heart attacks before and had new ones, had them within 24 hours of smoking pot. Totally disregarding that if they had another heart attack at all, the likelyhood of it being within 24 hours of smoking was high, even if there is no connection.
It wasn't even a large difference, it was a small anomaly from a small population.
In the end, the result wasn't worth reporting, much less a headline, but reporting it like they did got their names int he paper and a big national headline.