Statistical significance

For nearly all types of studies, a mathematical analysis must be made to see if the results are meaningful. It is always possible that an apparent result can be just due to chance. For example, if you flip a coin 20 times, and it comes up heads 14 times, does these mean that the coin is biased? Probably not. But if you flip it 4,000 times and it comes up heads 3,500 times, it probably is a trick coin.

Likewise, if a study only enrolls 20 people, the results might be due to chance alone. But results seen in larger studies are more likely to mean something.

Researchers use various statistical methods to analyze the outcome of a study and determine whether the results are meaningful. This analysis is called a test for statistical significance. You can't draw any conclusions from a study if the results are not statistically significant.