Statistically significant vs. actually big: what effect size means

There are two different questions hiding inside every study, and headlines love to blur them. The first is: is this effect real, or could it be a fluke? The second is: if it is real, is it big enough to care about? Statistical significance answers the first. Effect size answers the second, and it is the one that usually matters most to you.

Here is why the distinction bites. With a large enough group of people, even a tiny, meaningless difference can come out "statistically significant." A pill might lower blood pressure by an amount so small no one would ever feel it, yet a study of fifty thousand people can flag that difference as real. Significant does not mean large. It means unlikely to be pure chance.

Effect size is the number that tells you how much. A drug that adds two weeks of life and a drug that adds two years of life can both be statistically significant, but they are wildly different in practice. When a study reports an effect size, or better yet a plain number like months gained or points lowered, you can judge for yourself whether it is a big deal.

So when a headline shouts that something is "significant," ask the follow-up question the headline skipped: how big is the effect? Look for the actual numbers, not just the word significant. A real but tiny effect is often not worth changing your life over, and knowing the difference is most of what it takes to read health news well.