Abstract
Dark-adapted observers were asked to use a 5-point rating scale in reporting whether they had seen a briefly exposed 1° square slightly off the fovea. The square, at constant luminance, was presented only on half the trials (s trials); nothing was presented on the remaining trials (n trials). The results were compared to two theories of absolute visual detection: (1) A two-state low-threshold theory according to which one fixed threshold exists, which, however, may be exceeded on an appreciable proportion of catch trials. (2) Statistical-decision theory, according to which (a) no fixed thresholds exist at all, but merely decision criteria that the observer can alter in accordance with instructions, and (b) the internal events arising on s and n trials may be described by two Gaussian distributions. The second theory received much stronger support from the data obtained in these experiments.
© 1963 Optical Society of America
Full Article | PDF ArticleMore Like This
H. Richard Blackwell
J. Opt. Soc. Am. 53(1) 129-160 (1963)
Mitsuo Ikeda
J. Opt. Soc. Am. 53(11) 1305-1313 (1963)
C. H. Graham
J. Opt. Soc. Am. 53(9) 1019-1025 (1963)