Given that the high-alcohol preferring lines generally completed more trials and also received more forced trials on the delayed side, actual consumption did not track perfectly with the main effect of Alcohol Preference on indifference points. Perhaps a better index of optimal behavior is the amount of sipper access time that mice earned on choice (nonforced) trials. To cancel the effect of differential total trials between Alcohol Preference groups, the choice sipper access time was divided by total trials completed to yield a measure of behavioral efficiency: that is, ability to gain access to the sipper tube.
