Effects of intertrial interval, orientation, and adaptation on orientation discrimination measured with a yes-no signal detection paradigm

This paper describes a performance measure of lateral dominance, the Target Tests, which provides separate assessments of manual speed and accuracy within a relatively simple paper-and-pencil format. The Target Tests were administered to 111 normal adults, and scores were examined with respect to handedness, familial left-handedness, and sex. The tests significantly differentiated between right- and left-handers. Target scores were substantially correlated with the Harris Tests of Lateral Dominance, suggesting adequate reliability and validity. The means, standard deviations, and ranges for the preferred and nonpreferred hands for both handedness groups were nearly identical. The Target Tests provide a convenient instrument for assessing motoric lateralization of speed and accuracy in clinical and research settings.     

A model for detection and recognition with signal uncertainty

A model for signal detectability suggested by Luce is extended to situations in which the observer is uncertain of some important characteristic of the signal, such as frequency. By making a single assumption concerning the observer's covert response behavior, two solutions are obtained corresponding to qualitatively different behavior. Decrements in detectability and in recognition with uncertainty are shown to be particular functions of discriminability and detectability of the stimuli in other situations. Relevant experimental data are considered.This work was supported by grant NSF-G5544 from the National Science Foundation. The author is indebted to R. Duncan Luce for critical discussion of the content and helpful suggestions concerning the form of this paper. Useful comments have been made by David M. Green, Francis W. Irwin, Roger M. Shepard, and W. P. Tanner, Jr.     

Metacontrast as measured under a signal detection model

Metacontrast is assessed both by percent accuracy and by the dichotomous sensitivity and criterion measures of a signal-detection model in a forced-choice detection task. The results showed a monotonic increase in sensitivity with increasing mask delay, and an abrupt shift of subjects' response criterion at an 80 ms interstimulus interval. The results are discussed in terms of the interpretational advantages of a signal-detection approach to metacontrast investigation.     

Change detection inflates confidence on a subsequent recognition task

A face viewed under good encoding conditions is more likely to be remembered than a face viewed under poor encoding conditions. In four experiments we investigated how encoding conditions affected confidence in recognising faces from line-ups. Participants performed a change detection task followed by a recognition task and then rated how confident they were in their recognition accuracy. In the first two experiments the same faces were repeated across trials. In the final two experiments novel faces were used on each trial. Target-present and target-absent line-ups were utilised. In each experiment participants had greater recognition confidence after change detection than after change blindness. The finding that change detection inflates confidence, even for inaccurate recognitions, indicates recognition certainty can be a product of perceived encoding conditions rather than authentic memory strength.     

Change detection inflates confidence on a subsequent recognition task

A face viewed under good encoding conditions is more likely to be remembered than a face viewed under poor encoding conditions. In four experiments we investigated how encoding conditions affected confidence in recognising faces from line-ups. Participants performed a change detection task followed by a recognition task and then rated how confident they were in their recognition accuracy. In the first two experiments the same faces were repeated across trials. In the final two experiments novel faces were used on each trial. Target-present and target-absent line-ups were utilised. In each experiment participants had greater recognition confidence after change detection than after change blindness. The finding that change detection inflates confidence, even for inaccurate recognitions, indicates recognition certainty can be a product of perceived encoding conditions rather than authentic memory strength.     

Operator behavior in a dynamic visual signal detection task

The effects of changes in signal-to-noise ratio (SIN) and feedback state on the operator’s detection behavior, particularly derived operator thresholds, were investigated using a discrete visual signal. It was found that: (1) the operator behaves as if he has established different thresholds as a function of S/N; (2) his decision thresholds are relatively independent of the presence or absence of feedback; and (3) the operator will change his threshold accordingly for a change from one S/N to another. A model of the operator’s behavior, which introduced the concept of “operator” noise. yielded predictions of detection and false alarm probabilities which were extremely close to the observed values.     

Comparisons of reaction time patterns in a sequential visual recognition task with simple geometric forms

Right isosceles triangles, which differed in size and orientation, were presented in a sequential visual recognition task. Subjects produced an identity match response if the first self-presented stimulus, now in memory, was identical to a sequentially presented stimulus and a nonidentity match response if the subsequently presented stimulus differed in either or both size and orientation. The pattern of nonidentity recognition reaction times produced was compared to predicted patterns deduced from several theories of form information processing. The pattern of responses for both group and individual data were better described by serial and parallel self-terminating models with probabilistic assumptions than they were by template or deterministic serial or parallel models. A second study, in which identity and nonidentity response bias was equated, found that the identity responses were significantly faster than nonidentity responses, supporting wholistic processing for the identity matches and, when combined with the nonidentity match results, supporting a dual process theory.     

Sex composition and group performance in a visual signal detection task

Forty-eight Ss (24 males, 24 females) were run in groups of 4 on 720 trials of a 16-alternative forced-choice visual signal detection task requiring both individual and group decisions. Four types of groups were formed: all male, all female, mixed-alternate (seated in the order male-female-male-female), and mixed-adjacent (seated in the order male-male-female-female). There were no differences in performance between all-male and all-female groups, but mixed-sex groups had poorer performance, with mixed-adjacent groups having significantly lower sensitivity than homogeneous-sex groups. The results were interpreted as indicating that heterogeneity of groups with regard to sex interacts with seating pattern in affecting performance, probably due to the formation of information coalitions between like-sex group members when the seating pattern encourages such coalitions.     

The effect of cueing in a visual signal detection task

In a signal detection task four subjects were presented with stimulus cards, 50% of which contained a capital O in one of the four quadrants. Fifty per cent of these Os were surrounded by a small rectangle. This small rectangle served as a simultaneous cue, indicating the region where the O could appear. The other 50% were surrounded by a big rectangle encompassing the whole field. Fifty per cent of the blank cards contained a small rectangle, the other 50% a big one. Exposure duration was set at 15 and 20 ms. The results indicated that subjects, for whom these exposure durations led to performance exceeding chance level, used the simultaneous cue for reducing the field to be searched, resulting in higher sensitivity (d') in the small rectangle condition. Results were interpreted in terms of increasing efficiency of search as a result of which a higher quality of the icon was available for making decisions about the presence of the O.     

Visual signal detection measured by event-related potentials

We compared the performance of 10 well-functioning university students who had experienced a mild head injury (MHI) an average of 6.4 years previously and 12 controls on a series of standard psychometric tests of attention, memory, and thinking and on a series of auditory oddball vigilance tasks to which we also took event-related potentials (ERPs). The MHI and Control groups performed equivalently on all the psychometric tasks and on self-report questionnaires of everyday memory and attention difficulties. The MHI group performed more slowly and with lower accuracy on only the most difficult of the oddball tasks, yet they showed substantially and significantly reduced P300 amplitudes and subsequent attentuation on all the oddball tasks, both easy and difficult. There were no alterations of N1, P2, and N2 components. These data suggest that despite excellent behavioral recovery, subtle information processing deficits involving attention nevertheless may persist long after the original injury and may not be apparent on a variety of standard psychometric measures.     

The effects of aging on reaction time in a signal detection task

The effects of aging on response time are examined in 2 simple signal detection tasks with young and older subjects (age 60 years and older). Older subjects were generally slower than young subjects, and standard Brinley plot analyses of response times showed typical results: slopes greater than 1 and (mostly) negative intercepts. R. Ratcliff, D. Spieler, and G. McKoon (2000) showed that the slopes of Brinley plots measure the relative standard deviations of the distributions of response times for older versus young subjects. Applying R. Ratcliff's (1978) diffusion model to fit the response times, their distributions, and response accuracy, it was found that the larger spread in older subjects' response times and their slowness relative to young subjects comes from a 50-ms slowing of the nondecision components of response time and more from conservative settings of response criteria.     

Use of markov-encoded sequential information in numerical signal detection

Twelve Ss made binary decisions with feedback on numbers from one of two normal distributions with equal variances and unequal means. Sequences of distribution choices corresponded to first-order two-state Markov processes with probabilities of change of state of p₁ = p₂ = .50, p₁ = p₂ = .75, and p₁ = p₂ = .25. Performance was best (in d’ terms) when p₁ = p₂ ≠ .50. First-order sequential response dependencies tended to mirror the first-order stimulus dependencies. Violations of a fixed cutoff point decision rule were concentrated in the region of the average critical point, with a bandwidth of about 1/2σ, in which violations were strikingly more frequent than would be expected if they had occurred randomly. These results imply that in this task Ss are using a criterion-band decision rule instead of a fixed cutoff point rule, and that they are basing decisions in the region of the criterion band on information extracted from the sequence of decisions presented to them. The average bandwidth is generally different from the optimum bandwidth used by an ideal O in combining the two sources of information.     

Choice behavior in a sequential decision-making task

Many day-to-day decisions involving risk are based on a person's memory for the past sequence of outcomes. The study of sequential decision making investigates how individuals use information about the past sequence of outcomes in order to make risky decisions. In this article, a random walk decision model called the confidence strength model is evaluated which was designed to describe the cognitive processes individuals use to combine sequential information provided by event patterns with payoff information in a sequential decision-making task. Two related experiments are reported which were designed to test various assumptions of the confidence strength model. The particular task used to study sequential decision making was the differential payoff paradigm in which the decision maker was required to predict which of two uncertain events would occur on each trial. Following each choice, feedback was presented in the form of differential monetary payoffs which were contingent on the alternative-event combination occurring on that trial. Payoffs were manipulated by varying the loss for incorrect decisions associated with each alternative, and sequential information was manipulated by constructing event sequences with a fixed distribution of run lengths within each block of trials. The response measures included the proportion of trials one alternative was chosen over another and also the average choice speed for each alternative. The main theoretical result was that the confidence strength model provided a good overall fit to the pattern of results for the choice proportion measure, and also accounted for a majority fo the findings obtained from the choice speed measures, although there were some minor exceptions. The confidence strength model provided a better account of the present results and previous research than several alternative choice models that were considered.     

A signal detection analysis of the recognition heuristic

The recognition heuristic uses a recognition decision to make an inference about an unknown variable in the world. Theories of recognition memory typically use a signal detection framework to predict this binary recognition decision. In this article, I integrate the recognition heuristic with signal detection theory to formally investigate how judges use their recognition memory to make inferences. The analysis reveals that false alarms and misses systematically influence the performance of the recognition heuristic. Furthermore, judges should adjust their recognition response criterion according to their experience with the environment to exploit the structure of information in it. Finally, the less-is-more effect is found to depend on the distribution of cue knowledge and judges’ sensitivity to the difference between experienced and novel items. Theoretical implications of this bridge between the recognition heuristic and models of recognition memory are discussed.     

The effect of sequential dependencies on some signal detection parameters

The values of various sensitivity and response criterion parameters were calculated under different experimental conditions for both detection and recognition tasks. The effects of feedback and stimulus probability were found to be consistent with previous studies. When the values of the parameters were computed depending on the nature of the immediately preceding trial, the criterion parameters were found to be dependent on the identity of the preceding response, while those measuring sensitivity were dependent on its correctness. The dependence of the sensitivity parameters could be explained as an artifact of statistical bias, while the effect on the criterion parameters appears genuine. The nature of the phenomenon of statistical bias is discussed.     

Acoustic interference in a recognition task

Phonetic or acoustic characteristics of stimuli in recall studies have consistently been shown to generate more interference over short intervals than semantic characteristics. We performed three experiments to determine whether this characteristic also applied to recognition processes over short intervals. Subjects were shown study lists of eight words and were asked to recognize which of those words recurred on subsequent eight-word lists. Delays varied from 0 to 100 s between study and test, with a rehearsal-preventing task inserted during the delays. Phonetic distractors generated greater interference than semantic distractors at all delay levels, and semantic distractors generated greater interference than random distractors. These distractor-type differences were not present in long-term recognition tasks. The deterioration in performance over time was considerably slower in these recognition experiments than in previous recall studies.     

Evidence for graded central processing resources in a sequential movement task

In the present experiment, we examined slowing of the individual key presses of a familiar keying sequence by four different versions of a concurrent tone counting task. This was done to determine whether the same cognitive processor that has previously been assumed by the dual processor model (DPM) to initiate familiar keying sequences and assist in their execution, is involved also in the central processes of a very different task (viz. identifying tones and counting target tones). The present results confirm this hypothesis. They also suggest that in this particular situation the central processing resources underlying the cognitive processor can be distributed across the central processes of different tasks in a graded manner, rather than that they continue to behave like a single, central processor that serially switches between the central processes of the concurrently performed tasks. We argue that the production of highly practiced movement sequences can be considered automatic in the sense that execution of familiar movement sequences can continue without cognitive control once they have been initiated.