Time reallocation in a multiresponse environment: Effects of restricting response classes.

The team worked with rats in an operant chamber. The rats could press four levers. Each lever turned on darkness or white noise for a short break.

The researchers first watched which lever each rat liked most. Then they locked that lever so it would not work. They wanted to see if any simple rule could predict where the rat would move its time.

They tried the test four times with each rat, taking the lever away and then giving it back. They used an ABAB design to be sure any change was real.

No single rule guessed right. Sometimes the rats spent more time on the next-best lever. Sometimes they spread time across the rest. Sometimes they just sat.

The study found no clear pattern. The authors say you must watch and measure; you cannot assume.

Reynolds et al. (1968) showed that pigeons follow a clean matching law when all choices stay open. Baer et al. (1984) shows the law breaks when you remove one choice. The two papers fit like before-and-after photos.

Peters et al. (2013) also took away a top response, but with humans in a DRO setup. Both studies agree: after you block the favorite move, behavior scatters in ways rules can’t catch.

McDowell et al. (2021) later built a new equation that folds in reinforcer size plus rate. Their model could one day replace the simple rules Baer et al. (1984) tested and found lacking.

If you block a client’s top behavior—hand-flaps, escape to the hallway, screen time—do not guess where the time will go. Probe first. Run a quick assessment phase with the item or room removed, tally what fills the gap, then build your intervention on those real data.