Fig 1.
(A) Example of stimuli in the numerosity and average item size comparison tasks. The set of stimuli was created with two different total field areas of ~7.6° and ~5.8° diameter. (B, C) The two stimuli were shown either in sequential or simultaneous presentation mode. In separate sessions, participants were asked to judge which array contained more dots or which one contained the dots with the larger average size.
Table 1.
Demographic data and neuropsychological assessment.
Fig 2.
Results from Experiment 1 where control subjects were tested with sequentially presented stimuli.
(A-B) Proportion of errors as a function of ratio of the attended dimension during numerical (A) and average size (B) judgments. Different lines show the error rate when participants were tested with congruent (solid line) or incongruent (dotted line) trials. (C-D) Psychometric functions for the control group for the number (C) and average size (D) tasks. Black curves fit the entire dataset while light and dark gray curves fit trials that are the small and the big, respectively, within the unattended dimension. Data in E show the average (dark big diamond) and single subjects’ PSE difference (light gray small circles) during numerosity (on the x axis) and average size (on the y axis) comparison when the dataset was split for the magnitude of the unattended dimension (small-big).
Fig 3.
Results of Experiment 2 where the control group was tested with simultaneous presentation.
Results show a similar pattern despite the change in presentation mode. Congruency effect and bias from the unattended dimension are evident in the proportion of errors and group average fits during the average size task, but not during the numerosity task.
Fig 4.
Results of Experiment 3 where the dyscalculic group was tested with simultaneous presentation.
Differently from the control group (Fig 3), a tendency for congruency effects in accuracy at the most difficult numerical ratios, and bias from the unattended dimension in the group average fits, are visible during numerosity judgments.
Fig 5.
Absolute size of interference effect from the unattended dimension (unsigned PSE bias) arising when subjects in the control (gray symbols) and dyscalculic (black symbols) group judged numerosity (x axis) or average item size (y axis). Small circles represent individual subjects’ biases, large diamonds represent the group average ± sem. Arrows refer to average data values.
Fig 6.
Correlation analysis–unsigned bias and JND.
Correlation between the unsigned PSE bias and the overall precision during average size (A) and numerosity (B) judgments. Gray and black circles represent participants of the control and dyscalculic group, respectively.
Fig 7.
Correlation analysis–unsigned bias and mental calculation skills.
Correlation between the unsigned PSE bias in the average size (A) and numerosity task (B) and mental calculation skills. Only size interference during numerical judgment significantly correlates with math abilities, even when the factors of group and inhibitory skills are partialled out.