Limits...
Prior and present evidence: how prior experience interacts with present information in a perceptual decision making task.

Karim M, Harris JA, Morley JW, Breakspear M - PLoS ONE (2012)

Bottom Line: Subjects perform better when the first stimulus lies between the second stimulus and the global mean of all stimuli on the judged dimension ("preferred" time-orders) compared to the alternative presentation order ("nonpreferred" time-orders).The time-order effect had a greater influence on task performance than two of the explicit factors-task difficulty and noise-but not context.It also affords valuable insights into basic mechanisms of information accumulation, storage, sensory weighting, and processing in neural circuits.

View Article: PubMed Central - PubMed

Affiliation: School of Psychiatry, Faculty of Medicine, University of New South Wales, Sydney, Australia. m.karim@unsw.edu.au

ABSTRACT

Background: Vibrotactile discrimination tasks have been used to examine decision making processes in the presence of perceptual uncertainty, induced by barely discernible frequency differences between paired stimuli or by the presence of embedded noise. One lesser known property of such tasks is that decisions made on a single trial may be biased by information from prior trials. An example is the time-order effect whereby the presentation order of paired stimuli may introduce differences in accuracy. Subjects perform better when the first stimulus lies between the second stimulus and the global mean of all stimuli on the judged dimension ("preferred" time-orders) compared to the alternative presentation order ("nonpreferred" time-orders). This has been conceptualised as a "drift" of the first stimulus representation towards the global mean of the stimulus-set (an internal standard). We describe the influence of prior information in relation to the more traditionally studied factors of interest in a classic discrimination task.

Methodology: Sixty subjects performed a vibrotactile discrimination task with different levels of uncertainty parametrically induced by increasing task difficulty, aperiodic stimulus noise, and changing the task instructions whilst maintaining identical stimulus properties (the "context").

Principal findings: The time-order effect had a greater influence on task performance than two of the explicit factors-task difficulty and noise-but not context. The influence of prior information increased with the distance of the first stimulus from the global mean, suggesting that the "drift" velocity of the first stimulus towards the global mean representation was greater for these trials.

Conclusions/significance: Awareness of the time-order effect and prior information in general is essential when studying perceptual decision making tasks. Implicit mechanisms may have a greater influence than the explicit factors under study. It also affords valuable insights into basic mechanisms of information accumulation, storage, sensory weighting, and processing in neural circuits.

Show MeSH
Accuracy (d′) and response time for factors task difficulty, noise and implicit factor distance.Vertical bars represent within-subject SEM. A. Top figure shows d′ values significantly decreased for with increasing task difficulty, presence of noise, and for the further distance trials. There was an interaction between noise and distance. B. Lower figure shows there were no significant main effects for response time.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3362626&req=5

pone-0037580-g006: Accuracy (d′) and response time for factors task difficulty, noise and implicit factor distance.Vertical bars represent within-subject SEM. A. Top figure shows d′ values significantly decreased for with increasing task difficulty, presence of noise, and for the further distance trials. There was an interaction between noise and distance. B. Lower figure shows there were no significant main effects for response time.

Mentions: A similar analysis with the same subset of participants (44 out of the total 60) was conducted on trials that contained aperiodic noise, allowing a three-way analysis of task difficulty (easy, medium), noise (regular, noisy) and distance (closer, further). This additional analysis showed that there was a significant interaction for accuracy between noise and distance{5a}. Subjects’ performance diminished more greatly between closer and further trials for regular vibrations compared to noisy vibrations (Table 5, Figure 6a). There was no significant difference in response time for noise{5b} (Figure 6b).


Prior and present evidence: how prior experience interacts with present information in a perceptual decision making task.

Karim M, Harris JA, Morley JW, Breakspear M - PLoS ONE (2012)

Accuracy (d′) and response time for factors task difficulty, noise and implicit factor distance.Vertical bars represent within-subject SEM. A. Top figure shows d′ values significantly decreased for with increasing task difficulty, presence of noise, and for the further distance trials. There was an interaction between noise and distance. B. Lower figure shows there were no significant main effects for response time.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3362626&req=5

pone-0037580-g006: Accuracy (d′) and response time for factors task difficulty, noise and implicit factor distance.Vertical bars represent within-subject SEM. A. Top figure shows d′ values significantly decreased for with increasing task difficulty, presence of noise, and for the further distance trials. There was an interaction between noise and distance. B. Lower figure shows there were no significant main effects for response time.
Mentions: A similar analysis with the same subset of participants (44 out of the total 60) was conducted on trials that contained aperiodic noise, allowing a three-way analysis of task difficulty (easy, medium), noise (regular, noisy) and distance (closer, further). This additional analysis showed that there was a significant interaction for accuracy between noise and distance{5a}. Subjects’ performance diminished more greatly between closer and further trials for regular vibrations compared to noisy vibrations (Table 5, Figure 6a). There was no significant difference in response time for noise{5b} (Figure 6b).

Bottom Line: Subjects perform better when the first stimulus lies between the second stimulus and the global mean of all stimuli on the judged dimension ("preferred" time-orders) compared to the alternative presentation order ("nonpreferred" time-orders).The time-order effect had a greater influence on task performance than two of the explicit factors-task difficulty and noise-but not context.It also affords valuable insights into basic mechanisms of information accumulation, storage, sensory weighting, and processing in neural circuits.

View Article: PubMed Central - PubMed

Affiliation: School of Psychiatry, Faculty of Medicine, University of New South Wales, Sydney, Australia. m.karim@unsw.edu.au

ABSTRACT

Background: Vibrotactile discrimination tasks have been used to examine decision making processes in the presence of perceptual uncertainty, induced by barely discernible frequency differences between paired stimuli or by the presence of embedded noise. One lesser known property of such tasks is that decisions made on a single trial may be biased by information from prior trials. An example is the time-order effect whereby the presentation order of paired stimuli may introduce differences in accuracy. Subjects perform better when the first stimulus lies between the second stimulus and the global mean of all stimuli on the judged dimension ("preferred" time-orders) compared to the alternative presentation order ("nonpreferred" time-orders). This has been conceptualised as a "drift" of the first stimulus representation towards the global mean of the stimulus-set (an internal standard). We describe the influence of prior information in relation to the more traditionally studied factors of interest in a classic discrimination task.

Methodology: Sixty subjects performed a vibrotactile discrimination task with different levels of uncertainty parametrically induced by increasing task difficulty, aperiodic stimulus noise, and changing the task instructions whilst maintaining identical stimulus properties (the "context").

Principal findings: The time-order effect had a greater influence on task performance than two of the explicit factors-task difficulty and noise-but not context. The influence of prior information increased with the distance of the first stimulus from the global mean, suggesting that the "drift" velocity of the first stimulus towards the global mean representation was greater for these trials.

Conclusions/significance: Awareness of the time-order effect and prior information in general is essential when studying perceptual decision making tasks. Implicit mechanisms may have a greater influence than the explicit factors under study. It also affords valuable insights into basic mechanisms of information accumulation, storage, sensory weighting, and processing in neural circuits.

Show MeSH