Vigilanssi-ilmiön mittaamisesta ja selitysmahdollisuuksista

Abstract

The purpose of this study was to compare several methods of measuring vigilance in one experimental situation. The purpose was also to compare hypotheses deduced on the grounds of expectancy-, arousal-, observing response- and signal detection theories. During the test which lasted 2 hours the quality of the signal, which was either transient or nontransient, and the observing responses, which the subjects did or did not do, were varied. The vigilance performance was measured by the following methods: detections, false detections, reaction times, rates of observing responses, observed attention, level of skin-conductance, number of GSR-reactions, length of GSR-amplitudes to detections, d' and β. The vigilance test was a visual performance, in which the subjects watched light moving in a circle from hole to hole. The signal was a jump over one hole. Signals were programmed according to Mackworths scale, which means that the average interval between signals was 2.5 minutes. The subjects were 40 soldiers. They were divided into four groups, according to the results achieved in personal test Nesi (one form of EPI) on scales: neuroticism, impulsiveness and sociability. The results were analysed mainly by two and three way analysis of variance with repeated measurings. Performance, measured by detections, observed attention, d' and level of GSR-conductance (transient signal) declined (p < .01) in the same way during the vigilance test. The same trend (p < .05) was also found in the rates of observing responses and GSR-amplitudes to detections but it was not as significant as the abovementioned. In false detections, reaction times, rates of GSR-reactions and level of skin-conductance (nontransient signal) no significant change occurred during the vigilance test. The curve of β was inverted U (p < .05). The most successful description of results was achieved by Stroh's IAF-signal detection theory. The older traditional theories of vigilance (expectancy-, arousal-, observing response- and signal detection theory) were not successful in interpreting the results because they described only some aspects of information flow in the vigilance test and that is why the results of this study did not verify logically the prognosis deduced from the traditional theories. In this study changes in observing mechanism seemed to explain most of the weakening in the subjects' performances. Differences between subjects' performances were large within groups and the differences increased on most scales during the test. Interpretations of differences are not found in this study. Signal duration was found to change the correlation between detections and reaction times, which during a transient signal was positive and during a nontransient signal negative. On the grounds of the flicker-fusion-test it was concluded that our vigilance test, which lasted two hours, strained subjects' visual channel, and decreased their sensitivity in discrimination. In addition, the subjective evaluations about physiological and psychological sensations after the vigilance test showed that the test strained the subjects surprisingly.