Type I Error

Seeing something that isn't there—concluding an effect or pattern exists when it doesn't. A Type I error (false positive) means your test found a 'signal' that was actually noise. In business, this is launching a product based on a false-positive A/B test, seeing a trend in random data, or attributing success to a strategy that was actually luck. The more tests you run, the more likely you are to find false positives.

When to use it

When interpreting any statistical test, especially with multiple comparisons; when data seems to confirm what you hoped to find (confirmation bias amplifies Type I errors); when making decisions based on small samples or short time periods; when evaluating research claims.

How it can help

The antidote to pattern-seeking overconfidence. Every time data 'shows' a result, ask: could this be a Type I error? How many tests did we run? (The more tests, the higher the false positive rate.) What's the base rate? Is the effect size meaningful or just statistically significant? The multiple comparisons problem means that if you test 20 hypotheses at p<0.05, you expect one false positive by pure chance.