It was previously found (Armus, 1986, 1988, 1990) that greater response effort requirements led to a greater percentage of short (< 1 s) interresponse times (lRTs). The explanation previously proposed (Armus, 1986) was based on the assumption that more effortful responses produced more salient response-produced feedback stimuli, and that these became more strongly conditioned to the lever pressing response than in the case of less effortful presses and less salient feedback stimuli. A series of four experiments focused on the possibility this relationship between effort and the percentage of short IRTs might have been an artifact of the effective reinforcement schedule resulting from higher response effort. Experiment 1 examined the percentage of short IRTs under two schedules of reinforcement: CRF and VR-4. Experiment 2 used the same basic procedure as Experiment 1, except that there was a delay period of 1 second after each reinforcement, in which presses were not counted towards the schedule. In Experiment 3, the percentage of short IRTs which occurred only after a reinforced press under CRF and VR-4 schedules was examined, and in Experiment 4, partial responses were recorded under two different effort levels-50 g (high effort) and 7 g (low effort). The results of these experiments suggest that it is possible that the "effort effect" found by Armus was indeed an artifact of what was functionally a partial reinforcement schedule resulting from higher effort conditions.
Miller, Wendy S. and Armus, Harvard L.
"Effort Effect: An Artifact of Effort-Produced Partial Responses?,"
The Psychological Record: Vol. 48
, Article 12.
Available at: http://opensiuc.lib.siu.edu/tpr/vol48/iss4/12