Human Systems Integration Division @ NASA Ames

Previous work examined only at the speed (latency) or the accuracy of responses in letter identification, trying to quantify the effects of contrast polarity (black on gray vs. white on gray) on letter identifiability. Scores combining speed and accuracy show a different interaction of the polarity effect with contrast than latency or accuracy scores alone, and generate larger statistical effects for the polarity effect. Signal detection theory has shown that, because of response biases, it is necessary for detection experiments to have false alarms to accurately estimate detectability when response biases are present. Speed-accuracy trade-off theory shows that non-zero error rates are necessary to estimate effective stimulus strength in the presence of changing strategies. Results show that a very simple measure of effective stimulus strength can effectively be used to compensate for observers changing response strategies trading off speed and accuracy.