This report outlines a series of studies investigating two general problems relating to instrumental conditioning procedures for children. The first series of studies analyze the assumptions made with regard to the reliability of measures of response strength, the effect of sampling various aspects of conditioning data in deriving measures of response strength, and the persistance of conditioning effects over time. The second set of data analyze the effect of reinforcing stimuli on changing serially dependent processes. The data show that in this procedure children adopt nonrandom patterns of responding. The particular pattern adopted by the child is a significant determinant of the magnitude of behavior change produced by reinforcing stimuli.