(A) Grand average ERPs, plotted separately for blocks with different ITIs, for controls (n = 23; blue, left) and dyslexics (n = 25; red, right). Timing of the two tones in each trial is denoted by the short black bars under the plots. Insets: P2 range (denoted in gray) enlarged. Dyslexics’ P2 area was similar for 6 and 9 s ITIs, whereas controls’ P2 was larger for the 9 s interval. (B) The decrease in P2 adaptation (solid lines, circles, left scale; estimated magnitude at asymptote minus fitted curve) and the decrease in contraction bias (dashed lines, triangles, right scale; reflecting implicit memory decay) as a function of ITI. Symbols denote groups’ means, and plotted curves were fitted to these means. (C) Groups’ means and SEMs of the individually fitted parameters (to an exponential decay) of both P2 adaptation (top) and behavioral contraction bias (bottom). In both, dyslexics differed from controls only in the estimated rate of decay (P2: p<0.0005; : p<0.05).