Ce. Sumpter et al., RESPONSE FORM, FORCE, AND NUMBER - EFFECTS ON CONCURRENT-SCHEDULE PERFORMANCE, Journal of the experimental analysis of behavior, 70(1), 1998, pp. 45-68
Six hens responded on concurrent variable-interval (key-peck) variable
-interval (door-push) schedules of reinforcement in which the second-o
rder (fixed-ratio) requirements on the alternatives (Experiment 1) or
the required door forces (Experiment 2) were varied. The key-peck and
door-push response (measured as fixed-ratio completion) and time data
were well described by the generalized matching law. However, the mani
pulations of fixed-ratio requirement and required response force diffe
red in their effects. The manipulations of fixed-ratio size affected t
he response and time measures differently, producing fairly constant,
multiplicative biases only in terms of response allocation. It was arg
ued that variations in fixed-ratio size necessarily change the One all
ocated to that response unit, and thus changes in time bias were not n
ecessarily a fundamental effect of changing the ratio. In contrast, th
e changes in response bias were a fundamental result of changes in rat
io size. The response-force manipulations produced similar bias shifts
in terms of response and time allocation, but they appeared to combin
e with relative reinforcement rate to affect choice interactively. Spe
cifically, behavior appeared to be biased towards the least effortful
(i.e., key-peck) response, but the increases in door force had a large
r effect on bias when the hens were making this response infrequently
(on a lean schedule). The different effects of the fixed-ratio and res
ponse-force manipulations on concurrent performance were partially acc
ounted for by the differing times required to complete each response u
nit under those manipulations, but this would not account for the inte
raction. The interaction would be consonant with increased response ef
fort decreasing the effective value of the associated reinforcement sc
hedule.