Publications

Refereed Journal Articles

Synopsis: The paper is about integration of social information in individuals. It demonstrates that individuals are able to profit from information about others’ decisions even if this information does not infer any new objective information about the problem. It is noteworthy that this type of social learning is not acknowledged in the standard economic models of social learning. However, information about others’ decisions may have the opposite effect if participants come to compare with each other in terms of short-term performance. The effect depends on the organization of the information flows and is explained by Construal Level Theory. 

Synopsis: The paper makes two contributions. First, it documents direct observation of cognitive innovation: a small portion of participants create a new learning procedure that allows them to approximate the optimal policy of a relatively complex dynamic decision problem. In line with the heuristic method, this learning procedure involves a simplification of the search space and the application of domain-general learning rules to this simplified space. A surprising finding is that even though about half of the participants produce the key feature of the learning procedureseveral steady states of the dynamical systemvery high monetary incentives are unable to increase the portion of participants who realize the full potential of the learning procedure. Second, the paper offers a task that can serve as a workhorse for experimental designs on how to promote cognitive innovation in organizations.

Synopsis: The paper demonstrates that a computerized goal-based decision elicitation format improves the performance of decision makers. The goal-based format offers a comprehensive set of dimensions in terms of which the problem may be contemplated and economizes on cognitive resources for simulating in the mind various tradeoffs in terms of these dimensions. Importantly, the goal-based format does not do any of the following: (1) provide information beyond what is computable on the basis of the task instructions using elementary algebra; (2) provide any information on what the “optimal” policy is; (3) nudge participants towards any specific policy. Instead, key to its effectiveness is that it prompts and enables participants to think integratively about safety and short-term profit goals. The goal-based format mimics more structured decision processes in organizations.

Synopsis: This paper adds another piece of evidence to the already overwhelming case against context-independent risk preferences. Our results reveal that the relative "riskiness" of individual decisions made by participants who had identified themselves as 'male' or 'female' depended on whether the decisions were taken with or without the information about two other participants' decisions (who jointly made up an "observation group" throughout the experiment). This happened although the information provided does not reveal any new information (beyond the information provided in the instructions) that is decision-relevant under standard risk-preference perspectives. The results support a problem-solving perspective to understand decisions under uncertainty.