Perceived knowledge and expectations: Essays on individual choice behavior

Author(s): Melo Ferreira, Iansa | Advisor(s): Charness, Gary | Abstract: This dissertation is composed of three Experimental papers on individual choice behavior. On the first I discuss decision making process within ambiguous settings and how ambiguity aversion and contrast effects can affect choice behavior. In this work two types of ambiguity are defined: "Subjective Ambiguity", where outcome probabilities are well defined but not known to the decision maker; and "Intrinsic Ambiguity", where decision maker's perception about outcome probabilities are subject to sudden changes. Using these two concepts separately, I performed an experiment to test how does the flow of one's own knowledge perception behaves. The results suggest distinct choice patterns for each ambiguity type. I find that intrinsic ambiguity tends to enhance recency bias. Also the distance between 2 events has a clear effect when the decisions involve subjectively ambiguous events, but not so if probabilities are uncertain. Finally, the existence of an intermediate event with a distinct bias from that of the baseline does not reduce the influence of that baseline over the individual's knowledge perception. On the second paper I present a laboratory experiment designed to shed light into the role of expectations on workers' reciprocal behavior when a cut in wages takes place. Previous literature has studied reciprocity without expectation formation, but if expectations enter individuals' loss-gain utility, they may affect perception and effort choices. Using fixed wages and productivity-dependent profits, I measure workers' responses to a cut in wages in situations where that cut was more expected, as well as when it was less expected. Results are consistent with the notion of reciprocity and also with the idea that expectations can influence workers' reciprocal behavior, so that workers effort varies less when wages were expected. Lastly, the third paper explores the effect of a movie trailer over the enjoyment of a short movie, using a lab experiment. The results show that the average enjoyment for was higher for the Control group than for the Treatment (Advertisement) group. In fact, on average, the difference between individuals' predicted and realized enjoyments (before and after watching the movie) was twice as much for the Control than for the Treatment group.