A Reappraisal of Artificial Signals as an Aid to a Visual Monitoring Task

Three groups of subjects performed a visual search and detection tusk on un inherently noisy display. One group was given ‘ real ’ signals only, in a random 01 of the trials, without knowledge of results (KOK). Tho second group was given ‘ artificial’ signals, identical with the real signals, on a further 005 of the trials, increasing the apparent signal rate by 30%. The third group was similarly given artificial signals, and also immediate KOR on the trials in which artificial signals occurred, The subjects with artificial signals detected on average 75% of tho real signals against 63% detected without, but also made 22% compared with 12% false alarms. This indicates a difference in responso criterion even greater than normatively expected for the difference in signal probabilities, but there was no difference at all in discrimination efficiency. No benefit was conferred by KOR, probably because with variable signals no single form could be learned to serve in their discrimination.