📊 統計・データ

p 値

ぴーち

帰無仮説が真であると仮定したときに、観測されたデータ以上に極端な結果が得られる確率。統計的有意性の判定に使われる。

1 分で読める

Definition and Interpretation

A p-value is the probability of observing data as extreme as (or more extreme than) what was actually observed, assuming the null hypothesis is true. It does not tell you the probability that the hypothesis is correct. A p-value of 0.03 means there is a 3% chance of seeing such results if there were truly no effect - not that there is a 97% chance the effect is real.

The 0.05 Threshold and Its Problems

The conventional threshold of p < 0.05 for "statistical significance" is arbitrary and widely criticized. It creates a binary pass/fail mentality where p = 0.049 is treated as meaningful and p = 0.051 is dismissed, despite the negligible practical difference.

Publication bias compounds this problem: studies with p < 0.05 get published while those with p > 0.05 remain in file drawers, distorting the scientific literature toward positive findings.

Common Misinterpretations

The most frequent errors include believing a small p-value means a large effect, equating statistical significance with practical importance, and thinking a non-significant result proves no effect exists. P-values are heavily influenced by sample size - with enough data, trivially small effects become "significant."

Relevance to Ranking Data

When research claims that one country "significantly" outperforms another on a metric, check whether the difference is practically meaningful, not just statistically detectable. A ranking difference backed by a tiny effect size and a barely significant p-value should not change your behavior. Always pair p-value awareness with effect size thinking for sound interpretation.

関連用語

関連記事

この用語解説は役に立ちましたか?