Relative allocation on concurrent schedules can depend on schedule parameters when behavioral parameters are constant.

The authors built a math model of two schedules running side by side.

They asked: can the way we split our time look different even when the real behavior stays the same?

The answer was yes—if the organism rarely switches, the numbers can fool you.

Low switch rates can make the richer schedule look even richer than it is.

So a 70-30 split on paper may not mean the learner truly prefers that side.

The schedule itself, not the child, can tilt the scoreboard.

Glenn (1988) later showed rats almost never switch when only the active counter grows.

That real-data picture lines up with the model’s warning: stay quiet, get lopsided scores.

Bromley et al. (1998) worked with adults with intellectual disability and preschool kids on fixed-ratio schedules.

They saw people move to the harder task when the pay was better—clear choice.

Their clean data work because their learners switched often; the 1987 paper says if switching drops, the same method could mislead.

Davison et al. (1995) found pigeons act more indifferent than the matching law predicts when ratios hit 27:1.

Both papers shout the same message: raw allocation numbers need a second look.

Before you call a 80-20 split a strong preference, count how many times the client switched sides.

If switches are rare, run a few probe sessions with easier change-over delays or richer switch prompts.

Only then trust the proportion as a true voice of preference.