statistical significance

Not random chance

Statistical significance is a measure indicating that a research result is unlikely to have occurred by chance alone.

When a finding is statistically significant, the probability of observing it if there were no real effect falls below a chosen threshold—typically 5%.

The term is central to hypothesis testing in science and social research: The study found a statistically significant link between sleep and memory.

Statistical significance doesn't necessarily mean the effect is large or practically important, only that it's probably real.

testing

p-value

Probability under null hypothesis

ANOVA · hypothesis testing · degrees of freedom · F-test · parametric · nonparametric

inference

sample size

Quantity of observations taken

replication · generalizability · precision · reliability · validity · reproducibility · relative risk

errors

false positive

Incorrect presence indication

beta · underpowered · overpowered

thresholds

cutoff

Point of termination

research

peer review

Expert evaluation before publication

preregistration · scientific rigor

Categories
Opposites
statistical outcome

statistical significancesampling error

Subset-population value difference

noise · false negative · type II error

methodological rigor

statistical significanceanecdotal evidence

Support from individual accounts

intuition · gut feeling · clinical judgment · expert opinion · eyeballing

Family

significance, significant, statistics, insignificant, significantly, statistic, statistical, insignificance, signifying

Familiarity
Top 500,000
Profile
Used worldwide · Timeless · Abstract · Rated G