van der Meer and Redish (2009) found that reward-responsive cells in ventral striatum tended to be activated at the final choice point (T4) during early learning. To assess whether gamma power showed a similar change at the final choice point, we plotted the z-scored power (across spatial bins) for early (first 10 min of recording sessions, black line) and late (last 10 min, grey line) segments separately (Figure 8). For gamma-50, there was no significant difference between early and late power at T4 (T4 point ±2 spatial bins, two-sample Student's t-test: t(810) = −1.09, p = 0.23). For gamma-80, there was an overall tendency for gamma-80 to be elevated during early learning (two-way ANOVA with early/late and location (regular intervals centered around S, T1, T2, T3, T4, and after T4) as factors; main effect of early/late: F(1) = 78.0, p < 10−10), but this increase was not distributed uniformly across the maze (two-way ANOVA, early/late and location interaction: F(5) = 18.38, p = 10−10), with the largest difference occurring at T4 (two-sample Student's t-test, early vs. late: t(810) = 10.4,
