Figure 7. Timing of first and second CRs in sequence co-varies from trial-to-trial.

(A) Two example trials showing eyelid CRs from subjects trained in ipsilateral sequence. Orange dots indicate times of different CR timing measures, specified on the left. (B) Similarly, responses on two example trials from subjects trained in contralateral sequence. (C) Latency to criterion of following versus previous CR in ipsilateral sequence training. Each dot represents a single trial, dotted black line shows the diagonal. Colored lines show a linear regression fit for corresponding pairs of CRs as in the legend. Colored distributions on the right show distribution of time-intervals between timing measures (CR lat. to crit. versus CR lat. to crit. for C) of corresponding CRs pairs on different trials. Distributions in grey show the same data with timing of first CR shuffled across trials (2000 repetitions). (D) Similar plot for latency to criterion of CRs in contralateral sequence. Here colors indicate different gap intervals, as in the legend. Panels on the right show distributions of timing between left and right CRs data on the left. (E) Same for CR onset time versus peak time of the previous CR in ipsilateral sequence. (F) Same for the time to half peak amplitude of right CR versus latency to criterion of the left CR in contralateral sequence. For panels (C–F) dots corresponding to example trials from panels (A–B) are shown in orange.

Figure 7—figure supplement 1. Co-variation of timing measures of CRs in sequence.

Heatmaps show Pearson correlation coefficient of timing of different portions of CRs in the sequence (see Materials and methods). (A) Panel shows results for contralateral sequence of CRs (Δ = 300 ms). Eyelid position values were normalized between zero and one, such that CR onset time corresponds to zero, CR peak time corresponds to one and intermediate CR values are in between. X axis corresponds to the left CR, Y axis – to the right CR, color indicates the value of correlation coefficient as stated in the legend. Grey dots indicate points that correspond to panels in Figure 7 (approximately for CR latency to criterion). For contralateral sequence, the highest correlation is observed between timing of the middle portion of left CR and early portion of the right CR. (B-C) A similar analysis for ipsilateral sequence, timing of first versus second CRs and second versus third CRs. Here the highest correlation is observed between timing of the late portion of first CR and both early and late portions of the second CR. Interestingly, the timing of close to peak values of the second CR has a high correlation with the whole profile of the third CR.