One: Almost all cutting edge science is uncertain.

Only after repeated studies, ideally using different methods, does the answer approach certainty. At that point, it goes into the textbooks and scientists move on to a new mystery. This makes accurate stories about "new discoveries" hard to sell to editors, who naïvely think science should always be certain.

As a result, there is great temptation to hype the findings, to make them seem more certain. Bad move, because the next study on the same question could come to the opposite conclusion.

Two: Not all forms of research are equally convincing.

Clinical scientists, for example, speak of "levels of evidence" ranging from case reports (weakest) to observational studies (moderately persuasive) to interventional studies such as the "gold standard" of randomized, double-blind, placebo-controlled clinical trials (very convincing, if well done).

Always ask what other sources of evidence bear on the question at hand and where the new study fits in the spectrum. Some questions can never be addressed using clinical trials, usually either for ethical or cost reasons.

Three: Good journalistic balance does not necessarily mean giving equal weight to both sides of a controversy.

That kind of balance is naïve because it ignores facts and can mislead listeners into thinking there is equally strong evidence on both sides. Try to get a sense of the overall weight of the evidence and make sure your story roughly matches the actual balance in the real world. As research on a question progresses—usually over years—the balance may tip from 50-50 to 90-10 and, in some cases, to 100-0.

Four: The latest study on a topic is not necessarily closer to the truth than the previous studies.

Five: Separating fact from opinion is a bedrock function of journalism.

This can be difficult when scientists present weak evidence with strong opinion. Try to understand the science well enough to know whether the facts warrant the opinion. If you're unfamiliar with the science, you need to find additional sources to ask. One possible clue to the reason for strong scientific opinion is the study's funding source.

Six: The reporter's goal should always be accuracy and fairness.

If you're not sure you're meeting this goal, you probably need to do more reporting.

Boyce Rensberger is director of the Knight Science Journalism Fellowships program at MIT and the co-director of the summer Science Journalism Program at the Marine Biological Laboratory in Woods Hole, Massachusetts.