PSY BEH 11A Lecture Notes - Lecture 20: Detection Theory, Psychophysics, Standard Deviation

• School: UC-Irvine
• Department: Psychology and Social Behavior
• Course: PSY BEH 11A
• Professor: Bruce Berg

Document Summary

Measure(cid:373)e(cid:374)t of hu(cid:373)a(cid:374) (cid:271)eha(cid:448)ior is a little tri(cid:272)kier . Allows us to become more abstract when thinking about the problem. The classic idea: stimulus into a black box and then you get a response. Because of inherent biological noise, human observers are invariably inefficient relative to an ideal observer. In order to model the perceptual/decision process, internal noise is added to the observation. We made one box into two boxes. Coming up with an internal representation, x: sensory stage = processing stage, pass it onto the decision stage. Report (cid:862)sig(cid:374)al(cid:863) if the i(cid:374)ter(cid:374)al respo(cid:374)se is greater tha(cid:374) the (cid:272)riterio(cid:374) It represents the separation between the two distributions relative to the standard deviation: the higher the d", the higher the se(cid:374)siti(cid:448)it(cid:455) D" (cid:272)a(cid:374) also (cid:271)e (cid:448)ie(cid:449)ed as a(cid:374) i(cid:374)de(cid:454) of the sig(cid:374)al-to-noise ratio: d" = (cid:894) s n) / ( s n) is the sig(cid:374)al stre(cid:374)gth a(cid:374)d is the (cid:374)oise.