After researchers collect empirical data from a predetermined sample size, they can follow up with statistical analyses to determine whether the differences they observed between groups or variables are meaningful or the result of chance alone.

For example, perhaps a researcher finds that undergraduate students who were asked to use gesture and emotional expression to act out a scene—the experimental group—remembered more of their lines than students who read them without using gesture and emotional expression—the control group.

Now she wants to know the likelihood that this difference occurred because the experimental manipulation affected participants’ memory for the lines, rather than because of random happenstance.

To accomplish this task, she needs to establish the p-value—the probability that the difference between the groups occurred by chance.

If she finds that the p-value is 0.05 or below—meaning there is a five percent or less possibility the result occurred by chance—then the difference between the groups is considered statistically significant by convention.

In other words, there is at least a 95 percent chance that participants in the experimental group remembered more lines because of the experimental manipulation—using gesture and emotional expression.

As a result, she can confidently accept the alternative or experimental hypothesis—that the experimental manipulation affected the results—and reject the null hypothesis—that the experimental manipulation had no effect.

In the end, if research findings are found to be statistically significant, the results are considered meaningful differences by the scientific community.