Factors influencing the ratio bias

Abstract

The ratio bias refers to the tendency of individuals to judge probabilities expressed as ratios of large numbers as more likely than equivalent or even higher probabilities expressed as a ratio of small numbers. For example, the ratio bias effect occurs when a lottery offering a 9/100 chance of winning is preferred over a lottery that offers a 1/10 chance of winning. Although previous empirical research has found evidence for the ratio bias, the exact conditions under which this effect occurs are still unclear and there is a lack of rigor in distinguishing the ratio bias from other similar effects. In this article, besides providing a comprehensive and integrative literature review, we present the results of an experiment in which we extend previous research on the ratio bias by comparing deviations both in favor of low- and high-number alternatives, as well as allowing for indifference. Results indicate that a systematic deviation in favor of high-number alternatives does exist, but that the ratio bias must be clearly distinguished from a general tendency to indicate indifference. Concerning characteristics of the problem and the decision maker, we find significant influences of probability levels involved (the ratio bias occurs more frequently for low probabilities), and of gender (the bias occurs more often among female subjects).