Limits...
Temporal Integration of Auditory Information Is Invariant to Temporal Grouping Cues(1,2,3).

Liu AS, Tsunada J, Gold JI, Cohen YE - eNeuro (2015)

Bottom Line: Auditory perception depends on the temporal structure of incoming acoustic stimuli.We designed a novel discrimination task that required human listeners to decide whether a sequence of tone bursts was increasing or decreasing in frequency.We manipulated temporal perceptual-grouping cues by changing the time interval between the tone bursts, which led to listeners hearing the sequences as a single sound for short intervals or discrete sounds for longer intervals.

View Article: PubMed Central - HTML - PubMed

Affiliation: Bioengineering Graduate Group.

ABSTRACT
Auditory perception depends on the temporal structure of incoming acoustic stimuli. Here, we examined whether a temporal manipulation that affects the perceptual grouping also affects the time dependence of decisions regarding those stimuli. We designed a novel discrimination task that required human listeners to decide whether a sequence of tone bursts was increasing or decreasing in frequency. We manipulated temporal perceptual-grouping cues by changing the time interval between the tone bursts, which led to listeners hearing the sequences as a single sound for short intervals or discrete sounds for longer intervals. Despite these strong perceptual differences, this manipulation did not affect the efficiency of how auditory information was integrated over time to form a decision. Instead, the grouping manipulation affected subjects' speed-accuracy trade-offs. These results indicate that the temporal dynamics of evidence accumulation for auditory perceptual decisions can be invariant to manipulations that affect the perceptual grouping of the evidence.

No MeSH data available.


Related in: MedlinePlus

LATER model fits to signal RT data. A, Distributions of signal RT from 0%-coherence trials for one subject are plotted on a reciprobit plot: reciprocal RT versus percentage of cumulative frequency on a probit scale (Reddi et al., 2003), separately per IBI. Best-fitting values of the bound height (B) and mean rate-of-rise (C) of the LATER model (see Results and Materials and Methods for details) are plotted as a function of IBI for each subject and coherence (black/gray lines and data points). The data in black indicate that the model fits were improved significantly by fitting the given parameter separately for each IBI condition (likelihood-ratio test, p < 0.01, Bonferroni-corrected for two parameters). Shaded lines/symbols indicate that the model fits were not improved significantly. Red data points/lines represent the median values across all conditions.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4596088&req=5

Figure 6: LATER model fits to signal RT data. A, Distributions of signal RT from 0%-coherence trials for one subject are plotted on a reciprobit plot: reciprocal RT versus percentage of cumulative frequency on a probit scale (Reddi et al., 2003), separately per IBI. Best-fitting values of the bound height (B) and mean rate-of-rise (C) of the LATER model (see Results and Materials and Methods for details) are plotted as a function of IBI for each subject and coherence (black/gray lines and data points). The data in black indicate that the model fits were improved significantly by fitting the given parameter separately for each IBI condition (likelihood-ratio test, p < 0.01, Bonferroni-corrected for two parameters). Shaded lines/symbols indicate that the model fits were not improved significantly. Red data points/lines represent the median values across all conditions.

Mentions: These results were supported by independent analyses of the signal-RT distributions. A useful way to assess possible changes in drift rate and/or bound height in a simple accumulate-to-bound framework is to use the LATER model (Carpenter and Williams, 1995; Reddi et al., 2003). According to this model, a decision variable rises linearly to a threshold (bound) in order to trigger a motor response. Assuming a fixed bound, but a noisy decision variable with a rate of rise that is normally distributed across trials, RT is distributed as an inverse Gaussian. This distribution can be plotted as a straight line on “reciprobit” axes (i.e., percent cumulative frequency on a probit scale vs the reciprocal of RT from 0%-coherence trials; Fig. 6A). Horizontal shifts of these lines imply changes in the mean rate-of-rise of the decision variable, whereas swivels about a fixed point at infinite RT imply changes in the bound height (Reddi et al., 2003). When we fit the LATER model to signal-RT data separately for each subject, coherence, and IBI (correct trials only), we found that increasing IBI caused systematic decreases in the bound (Kruskal−Wallis test for H0: equal median values per IBI, across subjects and coherences, p < 0.001g; 34 of 36 individual subject−coherence pairs had a significant dependence of bound height on IBI, all of which had a lower bound for the longest vs the shortest IBI, p < 0.01h, likelihood-ratio test, Bonferroni-corrected for two parameters; Fig. 6B). In contrast, increasing IBI did not cause a systematic change in the best-fitting mean rate-of-rise (Kruskal−Wallis test, p > 0.05i; 13 of 36 individual subject−coherence pairs had a significant dependence of rate-of-rise on IBI, of which seven showed an increasing rate-of-rise and six showed a decreasing rate-of-rise; p < 0.01j, likelihood-ratio test, Bonferroni-corrected for two parameters; Fig. 6C).


Temporal Integration of Auditory Information Is Invariant to Temporal Grouping Cues(1,2,3).

Liu AS, Tsunada J, Gold JI, Cohen YE - eNeuro (2015)

LATER model fits to signal RT data. A, Distributions of signal RT from 0%-coherence trials for one subject are plotted on a reciprobit plot: reciprocal RT versus percentage of cumulative frequency on a probit scale (Reddi et al., 2003), separately per IBI. Best-fitting values of the bound height (B) and mean rate-of-rise (C) of the LATER model (see Results and Materials and Methods for details) are plotted as a function of IBI for each subject and coherence (black/gray lines and data points). The data in black indicate that the model fits were improved significantly by fitting the given parameter separately for each IBI condition (likelihood-ratio test, p < 0.01, Bonferroni-corrected for two parameters). Shaded lines/symbols indicate that the model fits were not improved significantly. Red data points/lines represent the median values across all conditions.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4596088&req=5

Figure 6: LATER model fits to signal RT data. A, Distributions of signal RT from 0%-coherence trials for one subject are plotted on a reciprobit plot: reciprocal RT versus percentage of cumulative frequency on a probit scale (Reddi et al., 2003), separately per IBI. Best-fitting values of the bound height (B) and mean rate-of-rise (C) of the LATER model (see Results and Materials and Methods for details) are plotted as a function of IBI for each subject and coherence (black/gray lines and data points). The data in black indicate that the model fits were improved significantly by fitting the given parameter separately for each IBI condition (likelihood-ratio test, p < 0.01, Bonferroni-corrected for two parameters). Shaded lines/symbols indicate that the model fits were not improved significantly. Red data points/lines represent the median values across all conditions.
Mentions: These results were supported by independent analyses of the signal-RT distributions. A useful way to assess possible changes in drift rate and/or bound height in a simple accumulate-to-bound framework is to use the LATER model (Carpenter and Williams, 1995; Reddi et al., 2003). According to this model, a decision variable rises linearly to a threshold (bound) in order to trigger a motor response. Assuming a fixed bound, but a noisy decision variable with a rate of rise that is normally distributed across trials, RT is distributed as an inverse Gaussian. This distribution can be plotted as a straight line on “reciprobit” axes (i.e., percent cumulative frequency on a probit scale vs the reciprocal of RT from 0%-coherence trials; Fig. 6A). Horizontal shifts of these lines imply changes in the mean rate-of-rise of the decision variable, whereas swivels about a fixed point at infinite RT imply changes in the bound height (Reddi et al., 2003). When we fit the LATER model to signal-RT data separately for each subject, coherence, and IBI (correct trials only), we found that increasing IBI caused systematic decreases in the bound (Kruskal−Wallis test for H0: equal median values per IBI, across subjects and coherences, p < 0.001g; 34 of 36 individual subject−coherence pairs had a significant dependence of bound height on IBI, all of which had a lower bound for the longest vs the shortest IBI, p < 0.01h, likelihood-ratio test, Bonferroni-corrected for two parameters; Fig. 6B). In contrast, increasing IBI did not cause a systematic change in the best-fitting mean rate-of-rise (Kruskal−Wallis test, p > 0.05i; 13 of 36 individual subject−coherence pairs had a significant dependence of rate-of-rise on IBI, of which seven showed an increasing rate-of-rise and six showed a decreasing rate-of-rise; p < 0.01j, likelihood-ratio test, Bonferroni-corrected for two parameters; Fig. 6C).

Bottom Line: Auditory perception depends on the temporal structure of incoming acoustic stimuli.We designed a novel discrimination task that required human listeners to decide whether a sequence of tone bursts was increasing or decreasing in frequency.We manipulated temporal perceptual-grouping cues by changing the time interval between the tone bursts, which led to listeners hearing the sequences as a single sound for short intervals or discrete sounds for longer intervals.

View Article: PubMed Central - HTML - PubMed

Affiliation: Bioengineering Graduate Group.

ABSTRACT
Auditory perception depends on the temporal structure of incoming acoustic stimuli. Here, we examined whether a temporal manipulation that affects the perceptual grouping also affects the time dependence of decisions regarding those stimuli. We designed a novel discrimination task that required human listeners to decide whether a sequence of tone bursts was increasing or decreasing in frequency. We manipulated temporal perceptual-grouping cues by changing the time interval between the tone bursts, which led to listeners hearing the sequences as a single sound for short intervals or discrete sounds for longer intervals. Despite these strong perceptual differences, this manipulation did not affect the efficiency of how auditory information was integrated over time to form a decision. Instead, the grouping manipulation affected subjects' speed-accuracy trade-offs. These results indicate that the temporal dynamics of evidence accumulation for auditory perceptual decisions can be invariant to manipulations that affect the perceptual grouping of the evidence.

No MeSH data available.


Related in: MedlinePlus