Research Interests and Working Papers

This page describes papers that have been presented at meetings or are not yet published but are available for distribution. Before citing any of the papers that you obtain here, please contact me at
{clemen -at- duke -dot- edu} to find out whether a newer version is available or has been published.

All files are in pdf format unless otherwise noted.

Table of Contents

Do prediction markets produce well calibrated probability forecasts?

by Lionel Page and Robert T. Clemen

Updated Sept 2011.

Abstract: Prediction markets have become the object of growing interest for their ability to extract and aggregate private information. In recent years, the ability of prediction market prices to accurately reflect the probabilities of underlying events has been brought into question. This paper presents new theoretical and empirical evidence on the forecasting ability of prediction markets. We develop a model that predicts that the time until expiration of a prediction market should negatively affect the accuracy of prices as a forecasting tool in the direction of a "favourite/longshot bias." That is, high-likelihood events are underpriced, and low-likelihood events are overpriced. We confirm this result using a large dataset of prediction market transaction prices. By studying the forecasting ability of market prices very precisely both overall and locally, we find that prediction markets are reasonably well calibrated when time to expiration is relatively short. Although calibration is worse for events farther in the future, an economic analysis shows that, when time value of money is considered, the miscalibration cannot be exploited to earn excess returns. In addition, political markets stand out as more poorly calibrated than other markets.

Click here to download

Psychological and organizational factors influencing decision process innovation: The role of perceived threat to managerial power

by Kelly E. See and Robert T. Clemen

September, 2005.

Abstract: Organizations often face complex choices involving uncertainty, trade-offs, and broad consequences, but responding to such situations in rational ways can be hampered by individual decision makers' cognitive limitations. The framework of decision analysis (DA) provides a unified collection of analytical decision-making tools and procedures that are designed to help managers cope with difficult decisions, yet little is known about what influences the ability of firms to innovate with respect to decision-making practices. This paper investigates factors that facilitate and impede adoption of decision process innovations . Integrating individual-level theories of technology acceptance and managerial innovation with organization-level theories of innovation, we present the results of a multilevel empirical survey of 160 senior managers from a variety of organizations. Our survey incorporates measures of individual psychological perceptions, organization structure, and environmental context. We find support for many of the variables that have previously been found to predict innovation, namely attitudes toward the innovation, organizational culture, degree of centralization, and concerns for legitimacy in the institutional environment. Furthermore, we examine a previously unexplored individual-level issue in innovation research, perceived threat to managerial value and control, and find that a key barrier to decision process innovation is the tendency for managers to perceive such innovations as threats to their own value, discretion, and control. This impediment to innovation is mitigated by highly formalized organizational structures, presumably because such structures are characterized by strict rules and hierarchy that are perceived to preserve authority and power.

Click here to download

Debiasing expert overconfidence: A Bayesian calibration model

by R. T. Clemen and K. C. Lichtendahl, Jr.

June, 2002. Presented at PSAM6, San Juan, Puerto Rico

Abstract: In a decision and risk analysis, experts may provide subjective probability distributions that encode their beliefs about future uncertain events. For continuous variables, experts often provide these judgments in the form of quantiles of the distribution (e.g., 5th, 50th, and 95th percentiles). Psychologists have shown, though, that such subjective distributions tend to be too narrow, representing overconfidence on the part of the expert. We propose an approach for modeling and debiasing expert overconfidence. Based on past performance data (previous assessments and realizations for a number of uncertain variables), and using Bayesian methods to update prior distributions on the model parameters, we show how our model can be used to debias expert probabilities. We develop and demonstrate both a single-expert model and a multiple-expert hierarchical model.

Click here to download

For other working papers related to decision analysis and education, check out: Decision Making in High Schools: A Curriculum-Development Project.