Radial Arm Maze Cognitive Testing. Figure 1 shows the results of radial arm maze cognitive testing. The number of errors is plotted versus the day of testing. Two-way ANOVA revealed a significant treatment effect in Tg 6D11 treated (n = 10) versus vehicle treated (n = 8) or murine IgG treated (n = 9) APP/PS1 Tg mice (p < 0.0001) with a Bonferroni post-hoc analysis showing no difference between Tg 6D11 treated and wild-type mice which were injected with either PBS alone (n = 8) or 6D11 (n = 9). APP/PS1 Tg non-treated mice and APP/PS1 mice treated with mouse IgG made significantly more errors than wild-type animals and 6D11 treated Tg mice (p < 0.01).