Two Os made identifications and certainty ratings for five Canadian traffic signs under a theory of signal detectability confidence-rating paradigm. The data showed a strategy from the theory of signal detection for the task as a whole but not for each separate target. Rather than referring to five separate signal and noise distributions, Os apparently used a single-decision axis for their judgments. This outcome suggests the need for clarification of the meaning of individual d' values obtained by a multi-target confidence-rating procedure, based on theory of signal detectability.
Get full access to this article
View all access options for this article.
References
1.
EdwardsA. L.Statistical methods for the behavioral sciences. New York: Rinehart, 1954.
2.
GreenD. M., & SwetsJ. A.Signal detection theory and psychophysics. New York: Wiley, 1966.
3.
LadanC. J.HeronR. M., & NelsonT. M.A signal detection evaluation of flat-curved marker performance. Perceptual and Motor Skills, 1974, 39, 355–358.
4.
Manual of Uniform Traffic Control Devices for Canada. (2nd ed.) Prepared by the Council on UTCDC, May, 1968.
5.
MarkowitzJ.DietrichC. W.LeesW. J., & FarmanM.An investigation of the design and performance of traffic control devices. Report No. 1726, 1968, Bolt, Beranek, & Newman, Contract No. CPR-11-5955.
6.
PetersonW. W.BirdsallT. G., & FoxW. C.The theory of signal detectability. Trans. IRE Professional Group on Information Theory, 1954, PGIT-4, 171–212.
