effect-size

How big a difference really is — not whether it exists, but whether it is large enough to matter, measured in units any two studies can share.

A p-value tells you whether a difference is probably real; an effect-size tells you whether it is worth caring about. Cohen's d (the mean difference divided by the standard deviation) is the common currency: d = 0.20 is small, 0.50 medium, 0.80 large. Two studies with different scales, samples, or measures can still compare their d-values directly — which is why meta-analyses average effect-sizes, not raw scores. The trap: a statistically significant result (p < .05) can have a tiny effect-size (d = 0.10), meaning the difference is real but too small to notice in practice. The honest question is never "is there an effect?" but "how large, and large enough for what?"