Sensitivity and Specificity

“Sensitivity” is from Latin “sensus” (feeling, perception); “specificity” from “species” (kind, sort), via “specific.” The pair was introduced to diagnostic testing by Jacob Yerushalmy in 1947 to evaluate screening reliability.

  1. noun phrase · /sen-sih-TIV-ih-tee and spes-ih-FIS-ih-tee/

    The two ways of scoring a test: how many real cases it catches, and how many non-cases it correctly clears.

    It puts the two failure modes of any test on the table at once, so “accurate” stops being a single vague word.

    A highly sensitive test rarely misses true cases but may over-flag; a highly specific one rarely false-alarms but may let cases slip through.

    Sensitivity is about avoiding false negatives and specificity about avoiding false positives; a test can be superb on one and poor on the other, which is why a single “percent accurate” figure is nearly useless.