Advertisement

Vigilance pp 679-703 | Cite as

Adaptation to Decreasing Signal Probability

  • D. Vickers
  • J. Leary
  • P. Barnes
Part of the NATO Conference Series book series (NATOCS, volume 3)

Abstract

Several recent studies of vigilance have explained the changes in an observer’s response to faint, improbable signals, appearing at unpredictable intervals, by postulating an increase in the observer’s criterion for a signal, as measured by the signal detection theory index β (Broadbent & Gregory, 1963b, 1965; Colquhoun & Baddeley, 1964, 1967; Williges, 1969). The present experiment was designed to investigate this criterion change more closely by progressively decreasing signal probability throughout a session.

Contrary to the usual findings in vigilance studies, the probability of correctly reporting a “signal” increased as a function of the length of the session, as did the probability of a “false alarm.” Conversely, the probability of both “missed signals” and “correct rejections” declined. A signal detection analysis revealed no change in sensitivity, but showed a significant decrease in β. While the present experiment studied effects of a decreasing, rather than a decreased, signal probability, and so differs from most types of vigilance tasks, the changes in subjective readiness usually postulated on a signal detection theory approach were the opposite to those observed. On the other hand, the present results appeared to be quite well accounted for by the hypothesis that observers respond to changes in the a priori probability of a signal response in such a way as to minimize any discrepancy between the local probability of making a signal response (estimated from a limited number of recent trials) and the cumulative probability of a signal response (reckoned over the entire sequence of trials). According to this adaptive model, the increase in β usually observed represents a return to stable performance after a very rapid decrease in β, made in response to a reduction in the local probability of a signal and a consequent discrepancy between the local probabilities of a signal response and its cumulative probability. The accuracy with which observers adapted to the progressive change in local probability encountered in the present experiment correlated significantly with their scores on the extraversion scale of the Eysenck Personality Inventory.

Keywords

False Alarm Test Phase Cumulative Probability Signal Probability Local Probability 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Plenum Press, New York 1977

Authors and Affiliations

  • D. Vickers
    • 1
  • J. Leary
  • P. Barnes
  1. 1.Department of PsychologyUniversity of AdelaideAdelaideAustralia

Personalised recommendations