Individuals are more likely to value skills that they are comparatively successful at (Festinger, 1954). At the same time, a person’s success on a particular domain is a consequence of his or her investment on that domain. This investment, in turn, can depend on how much the individual values that domain. That is, with social information, a rational choice must be based not only on the perceived decisions (on education, investment, entertainment, etc.) but also an unobservable selective valuation process in which one assigns a subjective value to the tasks/domains to optimize the observable choices. Studying how differences in values affect the behavior of an individual in tasks has important implications for our understanding of how individuals can optimize their choices and strategies in a society with competitions and social comparisons. In this study, we aim to design an innovative theory-driven experiment to investigate a new theory on social comparisons and estimate the effects of the selective valuation process on inequality, broadly defined. We will establish a new model to capture the characteristics of the selective valuation process in decision-making and derive theoretical predictions. We will then empirically explore the research questions in a controlled laboratory environment.