Type II error
Also known as: False negative, Beta error
Missing a real effect — failing to detect what's actually there.
What it means
A Type II error is the failure to reject a false null hypothesis, concluding there is no effect when one truly exists. Its probability is denoted beta, and one minus beta is statistical power — the chance of catching a real effect. Type II errors are driven mainly by small samples, small true effects, and noisy measurement, and they are rampant in underpowered literatures. Because Type I and Type II error rates trade off, sound design sets the significance level and then secures enough power to keep false negatives acceptably low.
Examples
A small pilot study finds 'no significant difference' between treatments that genuinely differ, simply because it lacked the participants to detect it.
A school evaluates new tutoring with an end-of-year test so noisy that pupils' scores swing wildly for unrelated reasons. The real gain is there; the measurement simply drowns it out.
A trial powered to detect a drug's benefit reports 'no evidence' of a rare side effect. That is not evidence the harm is absent — the study could never have seen it.
First described in Jerzy Neyman & Egon Pearson (1928, 1933).