Figure 5.

KO mice failed to discriminate between levers based on reward density. A, Two WT mice were excluded based on contingency-independent bias (>80% responses allocated to 1 pellet lever by Session 4 and >2 SDs from the group mean by day 8). B, Performance profile of KO mice prevented exclusions based on bias criterion. Each line indicates a single subject. C, KO mice (n = 10) did not acquire preference for the nutritionally profitable lever (3 pellet) as did most WT mice (n = 8). D, KO mice completed significantly fewer training/no choice trials (Block 1). E, KO mice and WT mice did not differ in choice trials completed (Block 2) or (F) cumulative rewards consumed (Sessions 1–8). Percent preference was calculated from responses made during choice trials (Block 2). Percent preference = no. of responses on 3 pellet lever/total no. of responses on both levers. *p < 0.05, **p < 0.01, ***p < 0.001 vs WT.