Limits...
Adaptive sampling of information in perceptual decision-making.

Cassey TC, Evens DR, Bogacz R, Marshall JA, Ludwig CJ - PLoS ONE (2013)

Bottom Line: Observers compared the direction of two random dot motion patterns that were triggered only when fixated.There were several differences between the optimal observer predictions and human behaviour.These differences point to a number of other factors, beyond the quality of the currently available sources of information, that influences the sampling strategy.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science, University of Bristol, Bristol, United Kingdom.

ABSTRACT
In many perceptual and cognitive decision-making problems, humans sample multiple noisy information sources serially, and integrate the sampled information to make an overall decision. We derive the optimal decision procedure for two-alternative choice tasks in which the different options are sampled one at a time, sources vary in the quality of the information they provide, and the available time is fixed. To maximize accuracy, the optimal observer allocates time to sampling different information sources in proportion to their noise levels. We tested human observers in a corresponding perceptual decision-making task. Observers compared the direction of two random dot motion patterns that were triggered only when fixated. Observers allocated more time to the noisier pattern, in a manner that correlated with their sensory uncertainty about the direction of the patterns. There were several differences between the optimal observer predictions and human behaviour. These differences point to a number of other factors, beyond the quality of the currently available sources of information, that influences the sampling strategy.

Show MeSH
Models of the decision process for two different two-alternative decision problems.(a) Decision model for a single source two-alternative forced choice decision. A frequently used stimulus in such tasks is the random dot kinematogram (RDK), which consists of a number of dots, only some of which move in a particular “signal” direction. Subjects are typically asked to decide in which of two directions the RDK is moving. With a single source of visual information the sensory data provide evidence for both alternatives. In motion discrimination tasks this evidence comes from neurons in area MT whose activity (firing rate) is tuned to respond to motion in a particular direction. At each moment in time, evidence from two populations of neurons is used to update a single decision variable. (b) Decision model proposed by Krajbich et al. [18] for a comparative two-alternative decision problem with two sources of visual stimuli. In their study, participants were asked to choose between two food items presented simultaneously in different locations on the screen. With each stimulus providing evidence about one particular alternative, the decision variable is no longer updated simultaneously by evidence for all available targets. Instead, as a target is fixated (target X in (b)) evidence supporting that target is generated and incorporated into the decision variable. During this time, the mechanisms that represent the evidence for the non-fixated target remain silent (shaded out branch for target Y in (b)).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3842256&req=5

pone-0078993-g001: Models of the decision process for two different two-alternative decision problems.(a) Decision model for a single source two-alternative forced choice decision. A frequently used stimulus in such tasks is the random dot kinematogram (RDK), which consists of a number of dots, only some of which move in a particular “signal” direction. Subjects are typically asked to decide in which of two directions the RDK is moving. With a single source of visual information the sensory data provide evidence for both alternatives. In motion discrimination tasks this evidence comes from neurons in area MT whose activity (firing rate) is tuned to respond to motion in a particular direction. At each moment in time, evidence from two populations of neurons is used to update a single decision variable. (b) Decision model proposed by Krajbich et al. [18] for a comparative two-alternative decision problem with two sources of visual stimuli. In their study, participants were asked to choose between two food items presented simultaneously in different locations on the screen. With each stimulus providing evidence about one particular alternative, the decision variable is no longer updated simultaneously by evidence for all available targets. Instead, as a target is fixated (target X in (b)) evidence supporting that target is generated and incorporated into the decision variable. During this time, the mechanisms that represent the evidence for the non-fixated target remain silent (shaded out branch for target Y in (b)).

Mentions: Much of the work on perceptual decision-making involves observers (humans or non-human primates) making a binary decision about a single stimulus [9], [10]. For instance, observers may view a noisy pattern and make a decision about the average direction of motion of the pattern (typically either to the left or right). In such tasks it has been observed that certain neurons in frontal and parietal areas integrate information from sensory areas that encode the evidence for alternative choices [11], [12]. It has been proposed [13]–[16] that these neural circuits implement the computation described by the diffusion model (Figure 1a) [17]. This model assumes that the brain computes a decision variable that corresponds to the integrated difference between inputs from sensory neurons selective for the two alternatives. The diffusion model has been shown, under certain assumptions, to be the continuum limit of sequentially computing the ratio of the likelihoods of the two alternatives [18], which can be used to form optimal decisions in fixed time and free-response decision problems [19], [20].


Adaptive sampling of information in perceptual decision-making.

Cassey TC, Evens DR, Bogacz R, Marshall JA, Ludwig CJ - PLoS ONE (2013)

Models of the decision process for two different two-alternative decision problems.(a) Decision model for a single source two-alternative forced choice decision. A frequently used stimulus in such tasks is the random dot kinematogram (RDK), which consists of a number of dots, only some of which move in a particular “signal” direction. Subjects are typically asked to decide in which of two directions the RDK is moving. With a single source of visual information the sensory data provide evidence for both alternatives. In motion discrimination tasks this evidence comes from neurons in area MT whose activity (firing rate) is tuned to respond to motion in a particular direction. At each moment in time, evidence from two populations of neurons is used to update a single decision variable. (b) Decision model proposed by Krajbich et al. [18] for a comparative two-alternative decision problem with two sources of visual stimuli. In their study, participants were asked to choose between two food items presented simultaneously in different locations on the screen. With each stimulus providing evidence about one particular alternative, the decision variable is no longer updated simultaneously by evidence for all available targets. Instead, as a target is fixated (target X in (b)) evidence supporting that target is generated and incorporated into the decision variable. During this time, the mechanisms that represent the evidence for the non-fixated target remain silent (shaded out branch for target Y in (b)).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3842256&req=5

pone-0078993-g001: Models of the decision process for two different two-alternative decision problems.(a) Decision model for a single source two-alternative forced choice decision. A frequently used stimulus in such tasks is the random dot kinematogram (RDK), which consists of a number of dots, only some of which move in a particular “signal” direction. Subjects are typically asked to decide in which of two directions the RDK is moving. With a single source of visual information the sensory data provide evidence for both alternatives. In motion discrimination tasks this evidence comes from neurons in area MT whose activity (firing rate) is tuned to respond to motion in a particular direction. At each moment in time, evidence from two populations of neurons is used to update a single decision variable. (b) Decision model proposed by Krajbich et al. [18] for a comparative two-alternative decision problem with two sources of visual stimuli. In their study, participants were asked to choose between two food items presented simultaneously in different locations on the screen. With each stimulus providing evidence about one particular alternative, the decision variable is no longer updated simultaneously by evidence for all available targets. Instead, as a target is fixated (target X in (b)) evidence supporting that target is generated and incorporated into the decision variable. During this time, the mechanisms that represent the evidence for the non-fixated target remain silent (shaded out branch for target Y in (b)).
Mentions: Much of the work on perceptual decision-making involves observers (humans or non-human primates) making a binary decision about a single stimulus [9], [10]. For instance, observers may view a noisy pattern and make a decision about the average direction of motion of the pattern (typically either to the left or right). In such tasks it has been observed that certain neurons in frontal and parietal areas integrate information from sensory areas that encode the evidence for alternative choices [11], [12]. It has been proposed [13]–[16] that these neural circuits implement the computation described by the diffusion model (Figure 1a) [17]. This model assumes that the brain computes a decision variable that corresponds to the integrated difference between inputs from sensory neurons selective for the two alternatives. The diffusion model has been shown, under certain assumptions, to be the continuum limit of sequentially computing the ratio of the likelihoods of the two alternatives [18], which can be used to form optimal decisions in fixed time and free-response decision problems [19], [20].

Bottom Line: Observers compared the direction of two random dot motion patterns that were triggered only when fixated.There were several differences between the optimal observer predictions and human behaviour.These differences point to a number of other factors, beyond the quality of the currently available sources of information, that influences the sampling strategy.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science, University of Bristol, Bristol, United Kingdom.

ABSTRACT
In many perceptual and cognitive decision-making problems, humans sample multiple noisy information sources serially, and integrate the sampled information to make an overall decision. We derive the optimal decision procedure for two-alternative choice tasks in which the different options are sampled one at a time, sources vary in the quality of the information they provide, and the available time is fixed. To maximize accuracy, the optimal observer allocates time to sampling different information sources in proportion to their noise levels. We tested human observers in a corresponding perceptual decision-making task. Observers compared the direction of two random dot motion patterns that were triggered only when fixated. Observers allocated more time to the noisier pattern, in a manner that correlated with their sensory uncertainty about the direction of the patterns. There were several differences between the optimal observer predictions and human behaviour. These differences point to a number of other factors, beyond the quality of the currently available sources of information, that influences the sampling strategy.

Show MeSH