highly significantp<0.001
We found that the presence of local extended or remote arm was, on average, associated with higher likelihoods of correct subsequent choices across animals ( Figure 4A ), and that when data from all animals were considered, there was a highly significant effect ( Figure 4B , mixed logistic regression; point estimate of fold change, (95% CI), p value: 1.50, (1.18~1.90), p<0.001, see also Figure S11B for analysis by theta cycle).