Detection models: Experimental tests with electrocutaneous stimuli

Contemporary psychophysical procedures were employed to determine the nature of the process for detecting rectangular electrical pulses delivered to the skin. Earlier experiments had shown that the somatosensory system is exceedingly responsive to small changes in the energy of an electrocutaneous stimulus, suggesting that sensory noise may be relatively unimportant in this modality. However, a rating scale study showed that both the high-threshold theory and Luce’s low-threshold theory cannot describe the detection process. The form of the ROC curve is consistent with a signal-detection model that assumes a greater variance for the signal plus noise distribution than for the noise distribution. Additional support for signal detection theory came from an experiment that examined the likelihood of a correct second choke when the first was incorrect in a four-alternative forced-choice paradigm. Therefore, the sensory excitation caused by internal noise can sometimes be greater than the excitation produced by an added stimulus. Electrocutaneous detection differs from auditory or visual in that d’ increases at a greater rate as stimulus intensity changes.