Premack gave the children two response alternatives, eating candy or playing a pinball machine, and determined which of these behaviors was more probable for each child.
The results were consistent with the Premack principle: only the children who preferred eating candy over playing pinball showed a reinforcement effect.
That is, only children who preferred playing pinball over eating candy showed a reinforcement effect.
Such observations led the team of Timberlake and Allison (1974) to propose the response deprivation hypothesis.
Reinforcement occurs only when the situation is set up so that access to the contingent response has been reduced relative to its baseline level.
In effect, the subject must subsequently increase responding to make up for the "deprivation" of the contingent response.