Within-session meal-size effects on induced drinking.

The team gave rats different meal sizes during a single session. They watched how much water the rats drank after each food pellet.

The goal was to see if bigger meals always create more drinking, or if the link curves.

Drinking rose, then fell, making an upside-down U. Medium-sized meals produced the most licks.

The jump came from rats starting more drinking bouts, not from longer sips.

Reid et al. (1983) saw a straight line: bigger meals always meant more drinking. The new curve does not erase the old data; it shows the earlier test used only small and large meals, missing the sweet spot in the middle.

Thompson et al. (1971) also found more drinking after larger pellets, but they looked at extinction bursts, not within-session dose. Their early hint aligns with the peak shown here.

Castilla et al. (2013) added food deprivation and shorter gaps between meals. Both tweaks pushed drinking higher, extending the meal-size curve into harder conditions.

Monkey work by Anger et al. (1976) found a similar upside-down U when they stretched the time between pellets, not the size. The same non-linear rule appears across species and parameters.

Reinforcement size can act like a drug dose: too little does nothing, too much shuts the behavior down. When you see collateral behavior spike and then drop, check if the reinforcer you are delivering has landed in the middle “too much” zone. Try trimming or adding a bite of food, a minute of screen time, or a token to find the sweet spot that keeps the client engaged without flooding them.