Fig. 5: Task had no impact on the time to peak decoding of hue or luminance polarity. | Nature Communications

Fig. 5: Task had no impact on the time to peak decoding of hue or luminance polarity.

From: Temporal dynamics of the neural representation of hue and luminance polarity

Fig. 5

a Tasks, 1-back luminance-polarity matching (left), 1-back hue matching task (right); note that the sequence of stimuli illustrated is the same for both tasks but the correct answers are different. b Left panel, Decoding luminance polarity (black trace) and hue (blue trace) using data from the luminance-polarity matching task. The data show the average of the identity decoding problems (see Fig. 1). Traces represent the average of 1000 bootstrapped time courses across eight sessions (two participants, four sessions each). Shading is SEM of the bootstrapped time courses. Vertical lines show the time to peak. The horizontal sequence of points above the x-axis show when decoding was greater than chance for four or more consecutive time bins (determined using within-subjects null decoding distribution and FDR corrected, see “Methods” section). Hue decoding peaked after luminance-polarity decoding (p < 0.001). Right panel, Decoding luminance polarity (black line) and hue (blue line) using data from the hue-matching task. Hue decoding peaked after luminance-polarity decoding (p < 0.001). Other conventions as for the left panel. c Left panel, Decoding luminance polarity using data from the luminance-polarity matching task (bold line) and the hue matching task (thin line). Shading represents the SEM. The time-to-peak for decoding luminance polarity identity was not impacted by task (p = 0.384). Right panel, Decoding hue using data from the hue matching task (bold line) and luminance-polarity matching task (thin line); the time-to-peak for decoding hue was not impacted by task (p = 0.427; p values reflect two-sided t-tests using a bootstrapping procedure, not adjusted for multiple comparisons). In all panels, the bolded traces show the decoding problems that align with the task. Luminance-polarity decoding was significant for a longer duration for data collected during the luminance-polarity matching task (left panel), while hue decoding was significant for a longer duration for data collected during the hue matching task (right panel).

Back to article page