Peak shifts across the population

Measurements of peak center-of-mass position are shown for the population of ‘17–17’ (C–H) and ‘17–18’ (I–N) pairs.  Scatterplots (C–E, I–K) compare center-of-mass positions in evoked conditions (R,C,NR) against baseline (‘Blank’).  Deviations from the diagonal indicate changes in relative spike timing between conditions.  Pairs with significant differences between evoked and baseline conditions are indicated by filled circles and 5-95% error bars (red lines), determined by bootstrap analyses computed from random cycles.  For clarity, plots are clipped at +/- 50 msec.  Above and right of each scatterplot are histograms of the timing relationships for baseline and evoked conditions, respectively.  Note that baseline interactions were strongly shifted in the surface-to-border direction (negative positions).  Lengths of lines at top-right corner of scatterplots correspond to 10 pairs per bin.  Total number of pairs does not coincide across the three conditions because not all pairs exhibited peaks with significant height under all conditions.  Bar plots (F–H, L–N) indicate number of pairs with positive, border–to–surface (e.g. R>B) and negative, surface–to–border (e.g. R<B) shifts in peak position, relative to baseline.  Dark bars indicate pairs with significant shifts.  Numbers above bars indicate significance of difference in number of pairs with positive versus negative shifts (two-tailed binomial test; significance values based on paired t-test are given in main text).  Significant shifts in the border–to–surface direction (R>B, C>B, N>B)  were more evident for 17–18 pairs.