A theorem that states that the distribution of means is approximately normal if the sample size is large enough (n 30), regardless of the underlying distribution of the original measurements.