ERIC Number: ED048765
Record Type: Non-Journal
Publication Date: 1970-Nov
Reference Count: N/A
Predicting Human Performance. I. Estimating the Probability of Visual Detection. Final Report.
Teichner, Warren H.; Krebs, Marjorie J.
This review is one in a series intended to develop methods which maximize the use of the existing scientific literature as a basis for predicting human performance. It is concerned with sensory performance in target detection, defined in terms of the "probability of detection" of a flash of light. Two conditions of detection are considered: where the target light has no background luminance (absolute threshold) and where some level of background luminance is present (contrast threshold). The literature on the absolute threshold for seeing and on the contrast threshold was examined and the data within each type was transformed into common units and collated across studies. In each case, Crozier's Law for the relation between the standard deviation and the threshold was found to hold, and a constant was developed. It now seems practical to estimate the probability of detection of any luminance or contrast given only a threshold value. The analysis also supports the general conclusion about the absolute threshold that for small visual areas and up to a critical duration, there is a reciprocity between duration and luminance, and that critical duration decreases with increasing retinal area. It did not support the generally held understanding that a similar reciprocity exists between luminance and area. (Author/MT)
Descriptors: Lighting, Literature Reviews, Performance Factors, Performance Tests, Prediction, Predictive Measurement, Predictive Validity, Probability, Research Reviews (Publications), Statistical Analysis, Visual Acuity, Visual Discrimination, Visual Perception
National Technical Information Service, Springfield, Virginia 22151 (AD-716 796, MF $.95, HC $3.00)
Publication Type: N/A
Education Level: N/A
Sponsor: Office of Naval Research, Washington, DC.
Authoring Institution: American Institutes for Research in the Behavioral Sciences, Silver Spring, MD.