1. See if the report has been published in a reputable scientific journal. Good journals are peer-reviewed, meaning that at least two independent scientists and the journal editor have examined the study. Peer review doesn't guarantee that a study is flawless, but it helps. By contrast, studies presented at scientific conferences usually are not reviewed as carefully, so the chance of error is much greater.


2. Look in the article for opposing points of view. Virtually no scientific study comes out without at least one expert disagreeing with it! In a courtroom, no jury member votes innocent or guilty without hearing from both the prosecutor and the defense. By the same token, you should make sure that you've heard the opposing side before deciding to believe any piece of research,.


3. Pay more attention to expert panels than to single studies. A report issued by an expert panel will have taken into account most of the relevant studies, and will have incorporated the insights of multiple experts. This doesn't guarantee that the report will be free of bias, but it helps.


4. Beware of "groundbreaking" studies. Studies that open up new areas of investigation are often less well done and more prone to error than the studies that follow. First studies of new treatments often show terrific results, only to be followed by much less dramatic findings in subsequent studies. A groundbreaking study is like a new high-tech start-up: It might be the next best thing, or it might be a flash in the pan.


5. Decide how the study's recommendations fit (or don't fit) into your entire lifestyle, and consider the alternatives. For example, a study showing that two hours a day of strenuous exercise reduces high blood pressure may not be meaningful to you if there's absolutely no chance that you will find the time; however, cutting down on salt might have the same effect, and you might actually do it.


6. See if you can tell exactly what the researchers did. Lots of news stories only report the conclusions of a study, not anything really about how the study was conducted. Use this acid test to see if you've really understood a study: Can you explain the study to a friend? Can you relate the steps the researchers took and how they interpreted what they found? And even when you're just focusing on the results of a study, be wary: In some cases, news coverage misrepresents them or takes them out of context.


7. See if the conclusions make sense to you. If a study finds that sleeping with the windows open increases job satisfaction for office employees (I just made this one up!), does the report give any indication of how those things might be connected? Do you buy it? Beware of claims that one thing causes another. Most studies can't really prove causality.


8. Bigger is usually better, but not always. Studies with fewer than 100 subjects are often too small to take all the relevant factors into account, such as variations in child age, sex, ethnicity, family situation, etc. Studies that are very large (1,000 or more) can usually handle these different factors, but often the results may be based on just a very few questions, because it's very expensive and difficult to interview large numbers of people in depth.


9. Make sure the study applies to you and your family. Many studies look at highly selected groups of children. In order to test a new headache medicine, for example, researchers might recruit children who were referred to a subspecialty headache program because their headaches had not responded to regular treatments. Is this really the drug you want to use for a minor headache?


10. Consider the size of the reported effects or risks. Just because something is statistically linked to an unfavorable outcome doesn't mean that the actual risk is big enough to worry about. For example, it is well known that strawberries can cause life-threatening allergic reactions, but the risk is tiny enough that most people don't think twice about serving them.

© Copyright 2004 The Dr. Spock Company. All Rights Reserved.