Fig. 5: Neural speech tracking response (n = 32) by level of interest.

A TRFs for high (black) vs. low (gray) interest segments, averaged across electrodes showing significant speech tracking. The shaded gray area indicates a time windows with a significant difference between conditions (p < 0.05, corrected). B Topographical map showing electrodes with significant differences in TRF amplitudes (p < 0.05, corrected; white circles) during the significant time windows. C Box plot showing the distribution of speech reconstruction accuracy (r values) of the decoding model for high vs. low interest conditions. Red lines indicate the median, boxes represent the IQR, whiskers extend to 1.5×IQR, and individual dots represent participants’ data points.