A mixed Experiment (1 or 2) × Sex × Alcohol Preference (High or Low) × Delay (0, 1, 2, 4, or 8) ANOVA was performed on choice sipper access time per trial. Main effects of Delay F(3.9,231.9) = 185.81, p < 0.001 and Alcohol Preference F(1,60) = 14.63, p < 0.001 were detected, but there were no interactions (ps > 0.16). This measure of behavioral efficiency corresponded well with indifference points, and is illustrated in Table 2. Again, this dependent variable suggests the costs, in terms of a loss of reinforcer access time, of the high preferring lines’ more impulsive strategy. As differences in drinking efficiency could cloud interpretations by allowing faster mice to get more reward in less sipper access time, an ANOVA was conducted to assess this. Mean drinking efficiencies were 0.0149 ± 0.0006, 0.0134 ± 0.0006, 0.0112 ± 0.0006, and 0.0121 ± 0.0009 ml/s sipper access ± SEM in HAP2, LAP2, HAP1, and HS/Ibg lines, respectively. No effect of Line was detected (p = 0.25).
