학술논문

Comparing magnitudes across dimensions: a univariate and multivariate approach
Document Type
Conference
Source
2016 International Workshop on Pattern Recognition in Neuroimaging (PRNI) Pattern Recognition in Neuroimaging (PRNI), 2016 International Workshop on. :1-4 Jun, 2016
Subject
Bioengineering
Signal Processing and Analysis
Visualization
Correlation
Symmetric matrices
Psychology
Magnetic resonance imaging
Analytical models
Neuroimaging
fMRI
representational similarity analysis
numerosity
length
magnitudes
Language
Abstract
Is there a common neural code underlying the representation of different quantity dimensions? In a high resolution fMRI protocol, we compared the activation evoked by the presentation of lines with different lengths, and sets of different numbers. We contrasted the results obtained with standard univariate analyses with a multivariate approach comparing the representational similarities within and between dimensions (i.e. few : short = many : long?). Together, our findings suggest that although the representations of number and size are co-localized in parietal cortex, these two quantity dimensions do not share a common representational code.