We first began the experiment with a simple sample test in which individuals had had two separate bets to make. Similar to the main urn experiment, participants were given the choice of two hypothetical tanks –each containing a certain proportion of green and red balls: green offering a payoff of $50 while a red draw would result in a $0 payoff. In the first iteration, Tank A contained a known 5-5 split while Tank B held an undisclosed number of red and green balls. Unsurprisingly, all but two individuals chose Tank A: illustrating the group’s strong preference for risk over ambiguity. It was a result we expected to extend into the second iteration, wherein Tank A’s proportion of red to green balls shifted to 6-4; despite the increased risk of drawing a red ball, participants should have still opted to select Tank A, as they had no feasible way of determining probability of either outcome for Tank B. The results, however, betrayed our expectations. Of the 15 participants, only 7 individuals chose Tank A –a plurality of individuals now decided that the uncertain option offered better prospects than the seemingly riskier option. This sample portion of the experiment was intended to establish a …show more content…
Risk is often seen as the biggest contributor to uncertainty, but it only relays a portion of the whole story. Ambiguity plays equally a large role in the way individuals make decision through a concealment of outcome probabilities. According to Ellsberg, aversion to ambiguity will cause individuals to decline the opportunity for higher payoffs if it means being able to select a bet with certain probabilities. Yet, the results from our experiment showed that many individuals winded up swapping their risk references at multiple points throughout the process. This may have reflected a failure on our part to communicate the instructions clearly, but it may have also been an issue of individuals being uninterested or all together apathetic about their decisions. In planning for the experiment, we attempted to avoid this by creating an aggregate payoff pot that would go to a single individual –rather than being split among multiple individuals, as has been the case for previous experiments. Another issue present hanging over the experiment was the relatively small sample size of the classroom. Although an inherent factor in the experiment, conducting any sort of statistical test to compare the results of one stage to another. If this experiment were to be ran again both an increased sample size and a stronger effort to communicate instructions would need to be implemented in order to