A Comparison of Sequential Sampling Models for Two-Choice Reaction Time.

The authors evaluated 4 sequential sampling models for 2-choice decisions--the Wiener diffusion, Ornstein-Uhlenbeck (OU) diffusion, accumulator, and Poisson counter models--by fitting them to the response time (RT) distributions and accuracy data from 3 experiments. Each of the models was augmented with assumptions of variability across trials in the rate of accumulation of evidence from stimuli, the values of response criteria, and the value of base RT across trials. Although there was substantial model mimicry, empirical conditions were identified under which the models make discriminably different predictions. The best accounts of the data were provided by the Wiener diffusion model, the OU model with small-to-moderate decay, and the accumulator model with long-tailed (exponential) distributions of criteria, although the last was unable to produce error RTs shorter than correct RTs. The relationship between these models and 3 recent, neurally inspired models was also examined. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

Timing and reaction time.

Because reaction time (RT) tasks are generally repetitive and temporally regular, participants may use timing strategies that affect response speed and accuracy. This hypothesis was tested in 3 serial choice RT experiments in which participants were presented with stimuli that sometimes arrived earlier or later than normal. RTs increased and errors decreased when stimuli came earlier than normal, and RTs decreased and errors increased when stimuli came later than normal. The results were consistent with an elaboration of R. Ratcliff's diffusion model (R. Ratcliff, 1978; R. Ratcliff & J. N. Rouder, 1998; R. Ratcliff, T. Van Zandt, & G. McKoon, 1999), supplemented by a hypothesis developed by D. Laming (1979a, l979b), according to which participants initiate stimulus sampling before the onset of the stimulus at a time governed by an internal timekeeper. The success of this model suggests that timing is used in the service of decision making. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

How normal and retarded individuals monitor and regulate speed and accuracy of responding in serial choice tasks.

Two experiments, with 34 mentally retarded (MR) 16–33 yr olds (WAIS IQs 43–76) and 34 CA-matched normal controls, investigated whether differences in the way that MR and non-MR monitor and regulate speed and accuracy of responding contribute to the slower and more variable performance of MR Ss on choice RT tasks. In Exp I, most MR Ss detected their errors as efficiently as controls, a finding that excludes the possibility that MR Ss do not monitor accuracy efficiently but achieve comparable levels of accuracy by consistently responding within slow RT bands that minimize likelihood of errors. Exp II showed that while a qualitatively similar trial-by-trial tracking mechanism mediated the performance of both groups, MR Ss were less efficient at constraining RTs within fast but safe bands. Increasing error probabilities at longer RTs suggested that momentary fluctuations in stimulus discriminability and/or attention affected RT variability in MR Ss. The RT patterns for various sequences of correct responses initiated and terminated by errors suggested that the effective past experience guiding trial-by-trial RT adjustments of MR Ss was short and inadequate and that this accounted for much of the remaining RT variability contributing to differences between MR and non-MR Ss. (53 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Theoretical interpretations of the speed and accuracy of positive and negative responses.

Examines different interpretations of the RT difference between positive and negative responses in 2-choice matching tasks. R. W. Proctor and colleagues (see PA, Vols 66:4888; 69:11757 and 11886; 71:11296; and 72:11336) hold that the difference represents a difference in processing between same and different judgments, whereas the author and M. J. Hacker (see PA, Vols 67:8976 and 69:11758) argue that the difference can be accounted for in terms of criteria settings. The ways in which several models (including sequential sampling models and mixture models) of choice RT and matching can account for this RT difference are examined, and one particular model, the author's diffusion model (see record 1982-02760-001), is fitted to data from 3 experiments. Results of these fits provide a clear interpretation of the RT difference in terms of criteria settings. It is concluded that interpretation of such positive–negative RT differences in the absence of a specific model is hazardous at best. (35 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Network interference and number-fact retrieval: Evidence from children's alphaplication.

[Correction Notice: An erratum for this article was reported in Vol 46(2) of Canadian Journal of Psychology Revue Canadienne de Psychologie (see record 2007-10803-001). Figure 1 was inadvertently omitted. Reprints of this article, available from the authors, will include this figure.] Investigated the origins of several phenomena of number-fact retrieval by asking 45 children (aged 8 yrs 2 mo to 10 yrs 7 mo) to memorize alphaplication facts (arithmetic-like memory items composed of letters instead of numbers). Ss' performance in the task showed paralleled aspects of simple arithmetic performance. There was a strong performance advantage for tie over nontie problems on both reaction time (RT) and errors. Specific errors frequently involved operand-related answers. Correct RTs and error rates across problems were closely linked. Correct answers to poorly learned problems tended to be the most common error responses. Performance was not as good for problems that were introduced later in the learning sequence. Results support the network-interference approach to number-fact retrieval of J. I. Campbell and D. J. Graham (see record 1986-16912-001). (French abstract) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A random-walk model of accuracy and reaction time applied to three experiments on pigeon visual discrimination.

A random-walk model of visual discrimination is described and applied to reaction time (RT) distributions from three discrete-trial experiments with pigeons. Experiment 1 was a two-choice hue discrimination task with multiple hues. Choice percentages changed with hue discriminability; RTs were shortest for the least and most discriminable stimuli. Experiments 2 and 3 used go/no-go hue discriminations. Blocks of sessions differed in reward probability associated with a variable red stimulus in Experiment 2 and with a constant green stimulus in Experiment 3. Changes in hue had a large effect on response percentage and a small effect on RT; changes in reward shifted RT distributions on the time axis. The “random-walk, pigeon” model applied to these data is closely related to Ratcliff's diffusion model (Ratcliff, 1978; Ratcliff & Rouder, 1998). Simulations showed that stimulus discriminability affected the speed with which evidence accumulated toward a response threshold, in line with comparable effects in human subjects. Reward probability affected bias, modeled as the amount of evidence needed to reach one threshold rather than the other. The effects of reward probability are novel, and their isolation from stimulus effects within the decision process can guide development of a broader model of discrimination. (PsycINFO Database Record (c) 2011 APA, all rights reserved)     

Missed targets are more frequent than false alarms: a model for error rates in visual search

In many visual search tasks, reaction times (RTs) for target detection are measured as a function of display size. The corresponding error rates are usually low but increase with increasing display size. Missed-target errors are more common than false alarms. In recent models of visual search, the error rates were attributed to a premature search termination and error rates increasing with display size were interpreted as indicating a speed-accuracy trade-off and an underestimation of search times per item (obtained from RT slopes). A model is described in which errors occur as a result of imperfect rather than incomplete search (i.e., it is assumed that there are task-specific probabilities of categorizing a target or a distractor incorrectly). Signal-detection theory is used to show that the observed error rate properties can be attributed to an optimized decision strategy. "Corrections" of RT data are thus questionable.     

"Network interference and number-fact retrieval: Evidence from children's alphaplication": Erratum.

Reports an error in "Network interference and number-fact retrieval: Evidence from children's alphaplication" by D. Jeffrey Graham and Jamie I. Campbell (Canadian Journal of Psychology Revue Canadienne de Psychologie, 1992[Mar], Vol 46[1], 65-91). Figure 1 was inadvertently omitted. Reprints of this article, available from the authors, will include this figure. (The following abstract of the original article appeared in record 1992-42555-001.) Investigated the origins of several phenomena of number-fact retrieval by asking 45 children (aged 8 yrs 2 mo to 10 yrs 7 mo) to memorize alphaplication facts (arithmetic-like memory items composed of letters instead of numbers). Ss' performance in the task showed paralleled aspects of simple arithmetic performance. There was a strong performance advantage for tie over nontie problems on both reaction time (RT) and errors. Specific errors frequently involved operand-related answers. Correct RTs and error rates across problems were closely linked. Correct answers to poorly learned problems tended to be the most common error responses. Performance was not as good for problems that were introduced later in the learning sequence. Results support the network-interference approach to number-fact retrieval of J. I. Campbell and D. J. Graham (see record 1986-16912-001). (French abstract) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Effects of truncation on reaction time analysis.

Many reaction time (RT) researchers truncate their data sets, excluding as spurious all RTs falling outside a prespecified range. Such truncation can introduce bias because extreme but valid RTs may be excluded. This article examines biasing effects of truncation under various assumptions about the underlying distributions of valid and spurious RTs. For the mean, median, standard deviation and skewness of RT, truncation bias is larger than some often-studied experimental effects. Truncation can also seriously distort linear relations between RT and an independent variable, additive RT patterns in factorial designs, and hazard functions, but it has little effect on statistical power. A promising maximum likelihood procedure for estimating properties of an untruncated distribution from a truncated sample is reported, and a set of procedures to control for truncation biases when testing hypotheses is appended. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Age and sex differences in reaction time in adulthood: Results from the United Kingdom Health and Lifestyle Survey": Correction.

Reports an error in "Age and sex differences in reaction time in adulthood: Results from the United Kingdom Health and Lifestyle Survey" by Geoff Der and Ian J. Deary (Psychology and Aging, 2006[Mar], Vol 21[1], 62-73). In the article, the link to supplemental material is incorrect. The correct supplemental link follows. Supplemental materials are available online at http://dx.doi.org/10.1037/0882-7974.21.1.62.supp. (The following abstract of the original article appeared in record 2006-03906-007.) Reaction times (RTs) slow and become more variable with age. Research samples are typically small, biased, and of restricted age range. Consequently, little is known about the precise pattern of change, whereas evidence for sex differences is equivocal. The authors reanalyzed data for 7,130 adult participants in the United Kingdom Health and Lifestyle Survey, originally reported by F. A. Huppert (1987). The authors modeled the age differences in simple and 4-choice reaction time means and variabilities and tested for sex differences. Simple RT shows little slowing until around 50, whereas choice RT slows throughout the adult age range. The aging of choice RT variability is a function of its mean and the error rate. There are significant sex differences, most notably for choice RT variability. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A Diffusion Model Analysis of the Effects of Aging in the Lexical-Decision Task.

The effects of aging on response time (RT) are examined in 2 lexical-decision experiments with young and older subjects (age 60-75). The results show that the older subjects were slower than the young subjects, but more accurate. R. Ratcliff s (1978) diffusion model provided a good account of RTs, their distributions, and response accuracy. The fits show an 80-100-ms slowing of the nondecision components of RT for older subjects relative to young subjects and more conservative decision criterion settings for older subjects than for young subjects. The rates of accumulation of evidence were not significantly different for older compared with young subjects (less than 2% and 5% higher for older subjects relative to young subjects in the 2 experiments). (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Is smoking automatic? Demands of smoking behavior on attentional resources.

The role of attention in the production of smoking behavior was investigated. Experienced and novice smokers were asked to perform a reaction time (RT) task under 4 conditions: while smoking (smoking), while mimicking all aspects of smoking except inhaling (pseudosmoking), while simply holding a cigarette (holding), and while not smoking (baseline). Experienced smokers' RTs increased during the pseudosmoking and holding conditions compared with baseline but did not differ between the smoking and baseline conditions, suggesting that attentional resources were not required for typical smoking behavior but were required to alter or inhibit smoking behavior. Novice smokers' RTs were slower during both the smoking and pseudosmoking conditions but not the holding condition, suggesting that novice smokers require the use of resources to smoke. Experiment 2 demonstrated that the differences in RT across conditions could not be explained by differences in urges. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

More on the speed and accuracy of positive and negative responses.

The account given by the diffusion model (Ratcliff, 1985) of the relation between positive and negative responses has been criticized by Proctor, who argues that the diffusion model provides no theory of criterion setting and that it violates psychophysical principles. In reply, I argue that psychophysical principles are not violated and that a quantitative theory of criterion placement is currently outside the scope of the diffusion model and many other models of memory. I point out that the strength of the diffusion model lies in its ability to account for aspects of the data other than just mean reaction time (RT), and in its generality across experimental paradigms. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Context effects in letter perception: Comparison of two theories.

This study showed that Elementary Perceiver and Memorizer (EPAM) can explain letter recognition phenomena earlier simulated by the connectionist Interactive Activation Model (IAM) of word perception by J. L. McClelland and D. E. Rumelhart (see record 1981-31825-001) and Rumelhart and McClelland (see record 1982-07091-001). EPAM, a model of learning and recognition in the form of a computer program, has previously successfully explained many aspects of learning and perception in a range of task environments (E. A. Feigenbaum and H. A. Simon, 1962). This study shows that the human data modeled by the IAM are at least as accurately simulated by EPAM. The fact that one model (EPAM) processes perceptions serially and the other (IAM) processes them in parallel plays no essential role in producing the observed context effects. Both connectionist and serial symbolic simulations can be designed to exhibit these effects. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

"Pessimistic noise effects": The perception of reaction times in noise.

Used a between-group design to examine the effect of loud noise on a 2-choice discrete reaction task and the judgments Ss made about self-produced RTs under these conditions. In Exp I, 70 Ss (aged 26–39 yrs) completed a 2-choice RT task and a concurrent RT rating task of speed. White noise was presented to Ss in the experimental groups. RTs were unexpectedly faster in noise, but Ss used more "slow" categories in describing them. The effect was not apparent when the same RTs were rerated a 2nd time under instructions that indicated that they were random time intervals. Also, the effect was not apparent when a new group of 14 undergraduates in Exp II rated the original RT data, again in noise. Exp III showed that when asked to predict average RTs produced by a hypothetical S in noise, 30 uniformed Ss (aged 26–39 yrs) predicted slow RTs. Results are considered in the light of the hypothesis that pessimistic expectancies about likely effects of noise may be a factor influencing performance. (French abstract) (12 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

There’s little return for attentional momentum.

Inhibition of return (IOR) refers to a delay in response time (RT) to targets appearing at a previously cued location. The prevailing view is that IOR reflects visual-motor inhibition. The “attentional momentum” account rejects this idea, and instead proposes that IOR reflects an automatic shift of attention away from the cued location resulting in slower RTs to targets presented there and speeded RTs to targets opposite the cue (an opposite facilitation effect or OFE). The drawback of this account is that J. J. Snyder, W. C. Schmidt, and A. Kingstone (2001) showed that there are few data to support the OFE, and no evidence that the OFE accounts for the IOR effect. Despite this evidence, several recent studies have promoted attentional momentum as a valid explanation for the IOR effect. Reanalysis of these recent studies and new data reveal, again, that IOR routinely occurs in the absence of the OFE, and when the OFE does occur, the IOR effect need not be present. This double dissociation invalidates attentional momentum as an explanation for the IOR effect. Extant data support an inhibitory explanation of the IOR effect. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Recalibrating the auditory system: A speed-accuracy analysis of intensity perception.

Recalibration in loudness perception refers to an adaptation-like change in relative responsiveness to auditory signals of different sound frequencies. Listening to relatively weak tones at one frequency and stronger tones at another makes the latter appear softer. The authors showed recalibration not only in magnitude estimates of loudness but also in simple response times (RTs) and choice RTs. RTs depend on sound intensity and may serve as surrogates for loudness. Most important, the speeded classification paradigm also provided measures of errors. RTs and errors can serve jointly to distinguish changes in sensitivity from changes in response criterion. The changes in choice RT under different recalibrating conditions were not accompanied by changes in error rates predicted by the speed-accuracy trade-off. These results lend support to the hypothesis that loudness recalibration does not result from shifting decisional criteria but instead reflects a change in the underlying representation of auditory intensity. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Neural nets for generalization and classification: Comment on Staddon and Reid (1990).

The neural net model of J. E. R. Staddon and A. K. Reid (see record 1991-03032-001) explains exponential and Gaussian generalization gradients in the same way as the diffusion model of R. N. Shepard (1958). The "cognitive" generalization theory of Shepard (see record 1988-28272-001), which also has been implemented as a connectionist network, goes beyond both of these models in accounting for classification learning. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Slowness and age: Speed-accuracy mechanisms.

Young and older adults' mechanisms of trial-by-trial control of accuracy and choice reaction times (RTs) were compared in 2,000 trials. With equal mean error rates, the older group's correct and error RT were longer, and their within-subject distribution was a linear function of the younger group's. Conditional accuracy functions (CAFs) were very similar in location and shape, with both groups achieving 95% accuracy at the same RT. Combining RT distributions with CAFs showed that the older group did not track their limits as often as the younger group, and they were more careful, having fewer very fast (near random) responses, more average speed responses in long error-free runs, and more slowing following an error. All participants were faster before an error and slower immediately after, but the older participants had coarser RT control. To compensate for this, the older participants produced slower responding to avoid the very fast, high-error part of the CAF. (PsycINFO Database Record (c) 2010 APA, all rights reserved)