The Influence of Presentation Order on Probability Judgments in a Representative Market Setting

Abstract

This field experiment investigated the contextual effect of presentation order on probability judgments supporting risky decisions in a representative setting: the betting markets of live horseraces. Participants were 54 contestants in a wagering tournament at a national thoroughbred racetrack. Prior to each of 11 races the expert gamblers recorded their estimated odds for all horses to win. Presentation order was manipulated by re-ordering the judgment-related information printed in the track’s program that had been provided for horses in an earlier race. Findings revealed the subjective probability estimates of most gamblers were not invariant with respect to presentation ordering. Probability judgments and risky choices were also associated with the predictive information provided for the gamble first processed during decision making. These results directly contradict calibration findings that probability judgments in this market are highly accurate and they are discussed in terms of the representative experimental design’s utility.