Science journalism is a challenging task, because you're the middle of an elaborate game of postman. And every step of the way, uncertainty gets removed because uncertainty doesn't get ratings...or requires mathematics to quantify.
See PHD Comics for the entire science news cycle. I was, approximately, the guy in the fedora with the notepad in the cycle.
The rule was that every equation you had in an article halved the readership; it had to be a really sexy article to get my editor to let me have ONE equation in the article. It only happened twice.
If you were lucky, you got a call from your features editor, and he sent you to a conference with some topics he wanted covered. Or he had you following up a press release at the University.
If you weren't lucky, you were hoovering up press releases, trying to arrange interviews, and writing the pieces 'on spec', hoping to interest an editor, do another round of fact checking, sell them, and keep the lights on.
Topics that sold well were anything dealing with alcohol, food, cars or computers, or anything talking about the End of the World.
The Internet has chopped the bottom out of the freelance science journalism market. Newspapers are dying, and quickly. Anyone with an Internet connection can get a Blogger account (hey, they let ME have one...) If you want to get it published, it has to be relevant and timely...which all sounds wonderful until it's time for fact checking. Most newspapers don't even have a science section any longer.
Eventually, I moved from science journalism to technical writing, and from technical writing to game design. And arguably, with my game designs, I've combined both tech writing and science journalism...except I don't have to bow to editor's whims, and I can put in all the equations I want! Muahahahah!