Synopsis of Rational Approaches to Learning and Development

In this dissertation, I study decision-making mechanisms in young children and non-human primates across multiple domains—including visual attention and overt choice—in order to discover the efficacy and limitations of rational cognitive theories. Rational cognitive theories posit that intelligent organisms should choose actions that maximize utility (e. Effective utility maximization requires an agent to generate accurate expectations about what will happen in the future—which, in turn, requires the agent to mentally model the world. Adults have a substantial amount of world experience upon which they can base such models. Very young children, however, have far less world experience, and thus—at least initially— possess no such advantage. They must sample observations from their environments in order to overcome their naïveté concerning the structure of those environments. They gradually infer complex representations and form abstract theories about the world through the sampling process. Sample by sample, they continuously build upon, update, and revise their theories. My dissertation investigates young children's choice behavior throughout development in order to understand the decision-making mechanisms that guide knowledge acquisition. The included studies test the efficacy and limitations of rational cognitive theories in order to better understand the decision-making mechanisms that guide early learning and behavior in children and monkeys. The studies span across multiple domains—including visual attention and overt choice. The results provide empirical evidence that learners rely on utility maximization both to build complex models of the world starting from very little knowledge and, more generally, to guide their decisions and behavior. Five experiments were conducted on infants, young children, and monkeys using visual and auditory stimuli presented in sequences of events. These experiments show that children are capable of rational decisions that optimize future utility and exhibit a U-shaped relationship between stimulus complexity and attention. Similarly, monkeys' attentional patterns are also guided by stimulus complexity. The dissertation opens by applying a rational cognitive approach to perhaps the most classic example of seemingly irrational behavior in young children—their poor track record in delay-of-gratification tasks, such as in the Stanford Marshmallow experiments (e.g., Mischel & Ebbesen, 1970). Though children apparently fail to maximize utility in such delay-of-gratification tasks, the cause of these apparent failures was not well understood (Chapter 2, " Rational Snacking: Young children's decision-making on the marshmallow task is moderated byyear-olds choose an immediately available low-value reward (e.g., one marshmallow) over one of high-value (e.g., two marshmallows) after a temporal delay (Mischel & Ebbesen, 1970). …