Sonar target detection as a decision process

Abstract

IN A VIGILANCE TASK WHICH SIMULATED SONAR TARGET DETECTION, 12 SS WERE REQUIRED TO REPORT THE OCCURRENCE OF 200-MSEC PULSES IN A 900 CPS TONE, WHICH WAS CONTINUOUSLY MODULATED IN AMPLITUDE BY A LOW-BANDWIDTH NOISE SOURCE. IN 3 45-MIN SESSIONS SS REPORTED SIGNALS ONLY WHEN QUITE CERTAIN OF THEIR IDENTIFICATION (SURE PROCEDURE); IN A FURTHER 3 SESSIONS THEY REPORTED ANY SIGNAL-LIKE SOUND (UNSURE PROCEDURE). A SUBSTANTIALLY HIGHER PERCENTAGE OF BOTH "WEAK" AND "STRONG" SIGNALS WAS DETECTED WITH THE UNSURE PROCEDURE. FALSE REPORT RATE WAS ALSO HIGHER WITH THE UNSURE PROCEDURE, BUT ANALYSIS SHOWED THAT THIS REFLECTED A CHANGE IN DECISION CRITERION RATHER THAN IN DISCRIMINATION EFFICIENCY. WITHIN-SESSION DECREMENT WAS SLIGHT, AND UNRELATED EITHER TO SIGNAL STRENGTH OR REPORTING PROCEDURE. THE RESULTS SUPPORT THE CONTENTION THAT DECISION PROCESSES PLAY A MAJOR ROLE IN DETERMINING PERFORMANCE AT THIS KIND OF TASK. (PsycInfo Database Record (c) 2025 APA, all rights reserved)