In signal detection theory, what does d' represent?

• A. The observer's bias toward responding 'signal present'
• B. The intensity of background noise
• C. The baseline level of arousal
• D. The observer's sensitivity to the signal

Answer: (D)

In signal detection theory, d' measures the observer's sensitivity to the signal—the ability to discriminate signal from noise. It quantifies how far apart the signal-present and noise-alone distributions are; a larger d' means clearer separation and better detection, while a small d' indicates the observer struggles to tell the two apart. Importantly, d' reflects discrimination capability independently of bias: the observer might favor saying “signal present” or “no signal” depending on their criterion, but that bias is separate from the true sensitivity captured by d'. In practice, d' is calculated from the hit rate and false-alarm rate as d' = z(hit rate) − z(false alarm rate), assuming Gaussian distributions with equal variance. So, d' specifically represents how sensitive the observer is to detecting the signal.