EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

2393. The Effect of Overconfidence on the Use of Recommender and Decision Support Systems

Invited abstract in session TC-7: Behavioural OR meets Information systems, stream Behavioural OR.

Tuesday, 12:30-14:00
Room: 1019 (building: 202)

Authors (first author is the speaker)

1. Ayşegül Engin
Business Aministration, University of Vienna
2. Rudolf Vetschera
Dept. of Business Decisions and Analytics, University of Vienna

Abstract

To benefit from digitization, decision support systems and recommender systems find wide business-to-business applications with the increasing digitization. To ensure that organizations benefit from implementing such systems, how employees interact with them as users is crucial. Certainly, modifications to system elements, such as design elements, can improve usability by mitigating the gap between what the user can work with and what the system offers. However, cognitive biases such as the hindsight bias and the overconfidence bias affect the usability of the system and cannot be mitigated through a system element modification. The present study focuses on the effect of overconfidence as a consequence of hindsight bias on the use of personalized recommender systems and a more general decision support system. This study aims to contribute to the literature on recommendation systems by showing that users benefit from recommender systems depending on their cognitive biases differently. Using an experimental approach, the study looks into the use of recommender systems and general decision support in repeated decisions with higher order structures. Because of the higher-order structure, the designed decision situation for this study entails uncertainty. To disentangle decision behavior from outcome, which is affected by the uncertainty posed by decision situation, the study uses a reinforcement learning model to describe decision-making behavior of users.

Keywords

Status: accepted

Back to the list of papers