Eike B. Kroll

Bargaining power does not matter when sharing losses - Experimental evidence of Inequality Aversion in the Nash bargaining game

While experimental research on social dilemmas focuses on the distribution of gains, we analyze social preferences in the case of losses. In this experimental study, the participants share a loss in a Nash bargaining game, but waiting time, instead of monetary losses, serves as an incentive. Further, we assume that the participants prefer less, rather than more, waiting time. Our experiment consists of four versions of the Nash bargaining game to allow for a comparison of four classical negotiation concepts (Nash, equal loss, equal gain, and Kalai–Smorodinsky) and an equal split of the overall waiting time. Our experimental evidence shows that an equal split better predicts the outcome of a Nash bargaining game involving losses than classical concepts do. Furthermore, the findings support that the participants resort to equal splits at the cost of their overall welfare.

Alternation Bias and Reduction in St. Petersburg Gambles: An Experimental Investigation

Reduction of compound lotteries is implicit both in the statement of the St. Petersburg Paradox and in its resolution by Expected Utility (EU).We report three real-money choice experiments between truncated compound-form St. Petersburg gambles and their reduced-form equivalents. The first tests for differences in elicited Certainty Equivalents. The second develops the distinction between ‘weak-form’ and ‘strong-form’ rejection of Reduction, as well as a novel experimental task that verifiably implements Vernon Smith’s dominance precept. The third experiment checks for robustness against range and increment manipulation. In all three experiments the null hypothesis of Reduction is rejected, with systematic deprecation of the compound form in favor of the reduced form. This is consistent with the predictions of alternation bias. Together these experiments offer evidence that the Reduction assumption may have limited descriptive validity in modelling St. Petersburg gambles, whether by EU or non-EU theories.

How does repetition of signals increase precision of numerical judgment

This paper investigates the processing of repeated complex information. The focus of this study is how precision of stated numerical estimates is influenced by repetition of the signal and information about the estimates of others. The key question is whether individuals use the law of large numbers in their estimates. In an experiment, participants are asked to estimate the number of points in a scatter plot, which is visible for a short time. The setting of the experiment allows for stating intervals and/or point estimates. Our analysis shows that the estimated interval gets smaller with each repetition of the signal, but the pattern does not follow the prediction of statistical models. The difference between their own estimates and information about the estimates of others does not lead to higher stated precision of the estimate, but does improve its average quality, i.e. the difference between answer and signal gets smaller.

How performance based payoffs influence estimates of complex information? An experimental study on quality and precision in estimation tasks

This paper investigates the processing of repeated complex information. The focus of this study is, whether additional information and the introduction of performance-based payoffs have an influence on judgement. Therefore, an experiment is designed to investigate the degree of precision and quality of interval estimates. The data shows that providing additional information decreases the precision of stated estimates, while it improves its quality. The same result is obtained, when performance-based payoffs are introduced to an otherwise hypothetical decision environment. That means, while the variation in the treatment variables increase the quality of the estimate, the precision as a reflection of the decision makers confidence is reduced.

Experimental evidence of context-dependent preferences in risk-free settings

This study investigates context effects in general and the compromise effect in particular. It is argued that earlier research in this area lacks realism, a shortcoming that is a major drawback to research conclusions and stated management implications. The importance of this issue is stressed by previous research showing that behavioral anomalies found in hypothetical experimental settings tend to be significantly reduced when real payoff mechanisms are introduced. Therefore, to validate the compromise effect, an enhanced design is presented with participants making binding purchase decisions in the laboratory. We find that the compromise effect holds for real purchase decisions, and therefore is validated, and is not an artificial effect in surveys on hypothetical buying decisions. While conclusions and implications for marketing managers, derived in previous work assume that context effects hold for real market decisions, the results created by this enhanced design close this gap in the literature.