What Is D-Prime in Signal Detection Theory?
d-prime is a sensitivity score from signal detection theory. It compares the standardized hit rate with the standardized false alarm rate, so it separates actual discrimination ability from a general tendency to answer yes or no.
How to Calculate D-Prime
The standard formula is d-prime = z(H) - z(FA), where H is hit rate and FA is false alarm rate. Hit rate is hits divided by signal-present trials. False alarm rate is false alarms divided by noise-only trials.
D-Prime Interpretation Table
Criterion c and Response Bias
Criterion c estimates response bias. A positive c usually means the observer is conservative and needs stronger evidence to say yes. A negative c usually means the observer is liberal and says yes more easily.
Why 0% and 100% Rates Need Correction
A hit rate of 100% or a false alarm rate of 0% produces an infinite z-score. The calculator applies correction only when a count-based rate is exactly 0% or 100%, which keeps d-prime finite without changing ordinary non-boundary rates.
Counts vs Rates Input
Use counts when you have the full signal detection table: hits, misses, false alarms, and correct rejections. Use rates when your paper, assignment, or dataset already gives hit rate and false alarm rate.