Maximum likelihood estimation of parameters of signal detection theory—A direct solution

Dorfman, Donald D.; Alf, Edward F.

doi:10.1007/bf02289677

Cited by 163 publications

(81 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is similar to Thurstone's (1928) solution without assuming unit correlation as he did. It yields values close to those of the maximum likelihood solution of Dorfman and Alf (1968) and is suitable for desk calculation.…”

Section: Notementioning

confidence: 66%

Stimulus intensity, catch trial effects, and the speed-accuracy tradeoff in reaction time: A variable criterion theory interptation

et al. 1974

View full text Add to dashboard Cite

Ss responded to a 1,000-Hz tone of 50, BO, or 100 dB. Catch trial conditions were none! blan~~ials, a red light, a noise, and an 1,BOO-Hz tone. Auditory catch. signals were o.f the same mteJ;tsltles. RT distributions in the first three conditions were well descnbed by a family of exponential growth functions dependent upon stimulus intensity and by the parameters of normal criterion distribu~ions dependent upon catch trial conditions and between-session variability. Perform~ce in the auditory catch trial conditions was not dependent upon the same set of sensory growth functions. Performa!Ice m these conditions was described by a two-dimensional analysis of information transmitted as a function of time and interpreted in terms of variable criterion theory: The speed-accuracy~rad~off in t~is situation appears to depend upon differential rates of growth of intensity and associative information and the criterion used in responding to this information.In the context of a decision theory approach to reaction time (RT), catch trial effects may be viewed as resulting from variation in the decision criterion. Grice (1972b) has supported this interpretation in a quantitative analysis of an experiment by LaBerge (1971). Grice's theory of response evocation assumes that, following stimulus onset, sensory information (V) grows in strength according to some continuous function, the rate of growth depending upon stimulus intensity. In conditioning, and presumably in choice situations, there is also an associative information component which grows in a somewhat slower fashion (Grice, 1972a). When this sensory (and, when applicable, associative) growth reaches the level of the S's decision criterion or reaction threshold (T), the response occurs. Thus, in RT the latency is determined by the rate of growth of V and by the level of the criterion. The criterion is assumed to depend upon the requirements of the task; set, attentional, and motivational factors; and individual differences. Under homogeneous conditions of performance, it is assumed to be a normally distributed random variable with mean, T, and standard deviation, a. Both of these parameters are influenced by experimental conditions and individual differences. The value determining the probability of response at a particular latency is E =

show abstract

Section: Notementioning

confidence: 66%

Stimulus intensity, catch trial effects, and the speed-accuracy tradeoff in reaction time: A variable criterion theory interptation

et al. 1974

View full text Add to dashboard Cite

show abstract

“…For a given listener and level profile, the hits and false alarms from each session constituted one point on the ROC curve. A maximum-likelihood procedure (Dorfman & Alf, 1968) was used for fitting a binormal model (Hanley, 1988). AUC and its variance were computed from the best-fitting estimates of slope and intercept (Metz, Herman, & Shen, 1998;Swets, 1979).…”

Section: Sensitivitymentioning

confidence: 99%

The temporal weighting of loudness: effects of the level profile

Oberfeld

Plank

2010

Atten Percept Psychophys

View full text Add to dashboard Cite

In four experiments, we studied the influence of the level profile of time-varying sounds on temporal perceptual weights for loudness. The sounds consisted of contiguous wideband noise segments on which independent random-level perturbations were imposed. Experiment 1 showed that in sounds with a flat level profile, the first segment receives the highest weight (primacy effect). If, however, a gradual increase in level (fade-in) was imposed on the first few segments, the temporal weights showed a delayed primacy effect: The first unattenuated segment received the highest weight, while the fade-in segments were virtually ignored. This pattern argues against a capture of attention to the onset as the origin of the primacy effect. Experiment 2 demonstrated that listeners adjust their temporal weights to the level profile on a trial-by-trial basis. Experiment 3 ruled out potentially inferior intensity resolution at lower levels as the cause of the delayed primacy effect. Experiment 4 showed that the weighting patterns cannot be explained by perceptual segmentation of the sounds into a variable and a stable part. The results are interpreted in terms of memory and attention processes. We demonstrate that the prediction of loudness can be improved significantly by allowing for nonuniform temporal weights.

show abstract

“…ROC curve methodology was originally developed for use in signal detection theory, but has since been utilized for a large number of pair-wise disease classification problems (e.g.,disease state -signal versus healthy state -noise), from psychometrics (Dorfman and Alf 1968) to radiology (Metz 1986). Recently, ROC methodology was extended to threeclass diagnostic problems by constructing a three-dimensional surface.…”

Section: Introductionmentioning

confidence: 99%

Assessing multiple-group diagnostic problems with multi-dimensional receiver operating characteristic surfaces: Application to proton MR Spectroscopy (MRS) in HIV-related neurological injury

2008

View full text Add to dashboard Cite

We present the multi-dimensional Receiver Operating Characteristic (ROC) surface, a plot of the true classification rates of tests based on levels of biological markers, for multi-group discrimination, as an extension of the ROC curve, commonly used in two-group diagnostic testing. The volume under this surface (VUS) is a global accuracy measure of a test to classify subjects in multiple groups and useful to detect trends in marker measurements. We used three-dimensional ROC surfaces, and associated VUS, to discriminate between HIV-negative (NEG), HIV-positive neurologically asymptomatic (NAS) subjects and patients with AIDS demential complex (ADC), using brain metabolites measured by proton MRS. These were ratios of markers of inflammation, Choline (Cho) and myoinositol (MI), and brain injury, N-acetyl aspartate (NAA), divided by Creatine (Cr), measured in the basal ganglia and the frontal white matter. Statistically significant trends were observed in the three groups with respect to MI/Cr (VUS=0.43; 95% confidence interval (CI) 0.33-0.53), Cho/Cr (0.36; 0.27-0.45) in the basal ganglia and NAA/Cr in the frontal white matter (FWM) (0.29; 0.20-0.38), suggesting a continuum of injury during the neurologically asymptomatic stage of HIV infection, particularly with respect to brain inflammation. Adjusting for age increased the combined classification accuracy of age and NAA/Cr (p=0.053). Pairwise comparisons suggested that neuronal damage associated with NAA/Cr decreases was mainly observed in individuals with ADC, raising issues of synergism between HIV infection and age and possible acceleration of neurological deterioration in an aging HIV-positive population. The threedimensional ROC surface and its associated VUS are useful for assessing marker accuracy, detecting data trends and offering insight in disease processes affecting multiple groups.

show abstract

Maximum likelihood estimation of parameters of signal detection theory—A direct solution

Cited by 163 publications

References 4 publications

Stimulus intensity, catch trial effects, and the speed-accuracy tradeoff in reaction time: A variable criterion theory interptation

Stimulus intensity, catch trial effects, and the speed-accuracy tradeoff in reaction time: A variable criterion theory interptation

The temporal weighting of loudness: effects of the level profile

Assessing multiple-group diagnostic problems with multi-dimensional receiver operating characteristic surfaces: Application to proton MR Spectroscopy (MRS) in HIV-related neurological injury

Contact Info

Product

Resources

About