Supplementary Figure 7: Differences between responses in naive and trained animals.

We resampled the distribution of responses in the naïve and trained animals one thousand times and computed the likelihood of observing differences between responses of naïve and trained animals, if the units are randomly assigned. The P values represent the likelihood of observing the naïve mean response in the resampled distribution of trained ΔnFR_(W-S) (n= 1000 resamples corresponding to the number of naïve neurons in each area, see Figure 6). The results and the p-values clearly indicate that the two distributions are statistically distinct.