Typically, p-values of less than 0.05 are considered to be statistically significant.
For treatment effect estimates, this means there is less than a 5% probability that an effect estimate as large or larger than the study results would have occurred by the play of chance, even if the treatment to which participants were allocated actually had no impact on the outcome.
We recommend against using the term “statistically significant” in reference to effect estimates for a number of reasons.
First, statistical significance is often confused with “ clinical significance” (importance), especially when “significant” is used rather than “statistically significant”;
Second, the cut-off for statistical significance (0.05) is arbitrary. Being told that the results are “not statistically significant” does not tell you whether they were informative or uninformative (inconclusive).
If you feel that this definition hasn't helped you to understand the term, click on our monkey to let us know.