What makes good science writing?

While working on my writing my dissertation now, I'm also involved in editing articles for the Embryo Project Encyclopedia. I think I enjoy this job because I'm working out both my science-brain and writing-brain. It can be a real challenge to transform an undergrad's piece of writing into publishable material, but it's also very rewarding.

In my program, Biology and Society, we are encouraged to take the Embryo Project writing and editing classes. Actually, you can read about it in the Huffington Post right now, where a former EP scholar shares her thoughts on the experience. In the editing class, we receive a lot of training on how to write clearly and logically about science. Some of the main lessons I've learned through the course and my own editing are:

• Present information in a logical order (i.e. big concept to detailed concept)
• Always explain uncommon words
• If you're going to use a technical word or acronym only once or twice, get rid of it and use a more common word
• Avoid nominalizations (making words into unnecessarily longer nouns)
• No passive voice

From what I've experienced, a lot of scientists struggle with everything except the first point, because the rest are regularly included, if not encouraged, in scientific writing such as journal articles. When you're writing for a scientific audience, it's quite like writing in a foreign language because you assume that your audience already has a grasp on that language. But when I write or edit for a public audience, I only assume that person has about a 9th grade education. Which is a good rule in general if you're speaking to an audience outside of your own field!

There are some good examples of science writing out there, such as the New York Times Science section, NPR's science section, Scientific American, and Wired. I was wondering what else my audience would recommend? And what, for you, makes science writing good?