Statistical Significance
The judgment that an observed result is unlikely under the null hypothesis, usually because the p-value falls below a preset threshold such as 0.05.
A result is statistically significant when its p-value falls below a significance level, called alpha, chosen before the analysis, most commonly 0.05. Significance means the observed effect would be unlikely if the null hypothesis were true, so the null is rejected. It is a statement about evidence against chance, not about size or importance: with a large sample, a trivial difference can be significant, and with a small sample, a real effect can fail to reach significance.
In a thesis, use significance language precisely. State your alpha level in the methodology chapter, report tests consistently, and never inflate "significant" into "important" or read "not significant" as "no effect." Pairing significance tests with effect sizes and confidence intervals gives examiners the full picture and protects your discussion chapter from overclaiming.
Related terms
Writing the thesis this term belongs to?
CiteDash takes a thesis from research question to a compiled document, with AI that cites only real papers and verifies every claim against its source.