Similarity and Consistency in Algorithm-Guided Exploration

50 Pages Posted: 17 Jan 2023

See all articles by Yongping Bao

Yongping Bao

University of Bremen

Ludwig Danwitz

University of Bremen

Fabian Dvorak

University of Konstanz - Faculty of Economics and Statistics; Eawag, Swiss Federal Institute of Aquatic Science and Technology

Sebastian Fehrler

IZA; University of Konstanz - Faculty of Economics and Statistics; Unversity of Bremen, SOCIUM; University of Zurich, Center for Comparative and International Studies (CIS)

Lars Hornuf

Technische Universität Dresden; CESifo (Center for Economic Studies and Ifo Institute)

Hsuan Yu Lin

University of Bremen

Bettina von Helversen

University of Bremen

Date Written: 2022

Abstract

Algorithm-based decision support systems play an increasingly important role in decisions involving exploration tasks, such as product searches, portfolio choices, and human resource procurement. These tasks often involve a trade-off between exploration and exploitation, which can be highly dependent on individual preferences. In an online experiment, we study whether the willingness of participants to follow the advice of a reinforcement learning algorithm depends on the fit between their own exploration preferences and the algorithm’s advice. We vary the weight that the algorithm places on exploration rather than exploitation, and model the participants’ decision-making processes using a learning model comparable to the algorithm’s. This allows us to measure the degree to which one’s willingness to accept the algorithm’s advice depends on the weight it places on exploration and on the similarity between the exploration tendencies of the algorithm and the participant. We find that the algorithm’s advice affects and improves participants’ choices in all treatments. However, the degree to which participants are willing to follow the advice depends heavily on the algorithm’s exploration tendency. Participants are more likely to follow an algorithm that is more exploitative than they are, possibly interpreting the algorithm’s relative consistency over time as a signal of expertise. Similarity between human choices and the algorithm’s recommendations does not increase humans’ willingness to follow the recommendations. Hence, our results suggest that the consistency of an algorithm’s recommendations over time is key to inducing people to follow algorithmic advice in exploration tasks.

Keywords: algorithms, decision support systems, recommender systems, advice-taking, multi-armed bandit, search, exploration-exploitation, cognitive modeling

JEL Classification: C910, D830

Suggested Citation

Bao, Yongping and Danwitz, Ludwig and Dvorak, Fabian and Fehrler, Sebastian and Hornuf, Lars and Lin, Hsuan Yu and Helversen, Bettina von, Similarity and Consistency in Algorithm-Guided Exploration (2022). CESifo Working Paper No. 10188, Available at SSRN: https://ssrn.com/abstract=4324211 or http://dx.doi.org/10.2139/ssrn.4324211

Yongping Bao (Contact Author)

University of Bremen ( email )

Universitaetsallee GW I
Bremen, D-28334
Germany

Ludwig Danwitz

University of Bremen ( email )

Universitaetsallee GW I
Bremen, D-28334
Germany

Fabian Dvorak

University of Konstanz - Faculty of Economics and Statistics ( email )

Universitaetsstr. 10
78457 Konstanz
Germany

Eawag, Swiss Federal Institute of Aquatic Science and Technology ( email )

Überlandstrasse 133
Dübendorf, 8600
Switzerland

Sebastian Fehrler

IZA ( email )

P.O. Box 7240
Bonn, D-53072
Germany

University of Konstanz - Faculty of Economics and Statistics ( email )

Universitaetsstr. 10
78457 Konstanz
Germany

Unversity of Bremen, SOCIUM ( email )

SOCIUM
Mary-Somerville-Str. 5
Bremen, +49
Germany

University of Zurich, Center for Comparative and International Studies (CIS) ( email )

Affolternstrasse 56
8050 Zurich
Switzerland

Lars Hornuf

Technische Universität Dresden ( email )

Dresden, 01307
Germany

HOME PAGE: http://www.hornuf.com

CESifo (Center for Economic Studies and Ifo Institute) ( email )

Poschinger Str. 5
Munich, DE-81679
Germany

Hsuan Yu Lin

University of Bremen ( email )

Universitaetsallee GW I
Bremen, D-28334
Germany

Bettina von Helversen

University of Bremen ( email )

Universitaetsallee GW I
Bremen, D-28334
Germany

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
106
Abstract Views
484
Rank
469,670
PlumX Metrics