Thesis (Ph. D.)--Princeton University, 1988. / Includes bibliographical references (leaves 169-171).

In two experiments, groups of children received interest money contingent upon their savings; yoked control subjects received identical interest irrespective of their savings. In the first experiment, interest was paid every 3 days by parents for the average savings of the period. In Experiment II, interest was paid on a daily basis and control subjects received instructions concerning the lack of a relationship between interest and their savings. In both experiments savings and magnitude of expenditures increased systematically across the research phases for contingent and non-contingent interest subjects alike. Neither interest nor subject wealth was found to relate to subject savings, or expenditures in Experiment I. Contingent-interest subjects' frequency of expenditures, during the interest and reversal phases was reduced relative to that of baseline and the expenditures of yoked controls. Additionally, an increase in the conditional probability of an expenditure at longer inter-expenditure intervals developed during the interest phase and was maintained into the reversal phase. The contingent interest procedures of Experiment II differentially reinforced longer times between expenditures without affecting savings. Control subjects who received instructions regarding the absence of a relation between savings and interest did not alter the time between their expenditures.