Exploration versus exploitation: A laboratory test of the single-agent exponential bandit model

Stanton Hudja; Daniel Woods

doi:10.1111/ecin.13164

Exploration versus exploitation: A laboratory test of the single-agent exponential bandit model

Stanton Hudja, Daniel Woods^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

22 Downloads (Pure)

Abstract

This paper analyzes how individuals resolve an exploration versus exploitation trade-off in a laboratory experiment. The experiment implements the single-agent exponential bandit model. We analyze how subjects respond to changes in the prior belief, safe action, and discount factor. We find that subjects respond in the predicted direction to these changes. However, we find that subjects under-respond to the prior belief, under-respond to the safe action, and typically explore less than predicted. Our results suggest that neither risk aversion nor the random termination probability are driving under-experimentation. Our results are consistent with subjects having incorrect beliefs about exploration.

Original language	English
Pages (from-to)	267-286
Number of pages	20
Journal	Economic Inquiry
Volume	62
Issue number	1
DOIs	https://doi.org/10.1111/ecin.13164
Publication status	Published - Jan 2024
Externally published	Yes

Bibliographical note

© 2023 The Authors. Economic Inquiry published by Wiley Periodicals LLC on behalf of Western Economic Association International. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

Keywords

bandits
continuous time
experimentation
exponential bandit model

Access to Document

10.1111/ecin.13164Licence: CC BY-NC

Publisher version (open access)Final published version, 1.09 MBLicence: CC BY-NC

Cite this

@article{85be15430ad34e4f9fc28c9b372c0665,

title = "Exploration versus exploitation: A laboratory test of the single-agent exponential bandit model",

abstract = "This paper analyzes how individuals resolve an exploration versus exploitation trade-off in a laboratory experiment. The experiment implements the single-agent exponential bandit model. We analyze how subjects respond to changes in the prior belief, safe action, and discount factor. We find that subjects respond in the predicted direction to these changes. However, we find that subjects under-respond to the prior belief, under-respond to the safe action, and typically explore less than predicted. Our results suggest that neither risk aversion nor the random termination probability are driving under-experimentation. Our results are consistent with subjects having incorrect beliefs about exploration.",

keywords = "bandits, continuous time, experimentation, exponential bandit model",

author = "Stanton Hudja and Daniel Woods",

note = "{\textcopyright} 2023 The Authors. Economic Inquiry published by Wiley Periodicals LLC on behalf of Western Economic Association International. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.",

year = "2024",

month = jan,

doi = "10.1111/ecin.13164",

language = "English",

volume = "62",

pages = "267--286",

journal = "Economic Inquiry",

issn = "0095-2583",

publisher = "Wiley-Blackwell, Wiley",

number = "1",

}

TY - JOUR

T1 - Exploration versus exploitation

T2 - A laboratory test of the single-agent exponential bandit model

AU - Hudja, Stanton

AU - Woods, Daniel

N1 - © 2023 The Authors. Economic Inquiry published by Wiley Periodicals LLC on behalf of Western Economic Association International. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

PY - 2024/1

Y1 - 2024/1

N2 - This paper analyzes how individuals resolve an exploration versus exploitation trade-off in a laboratory experiment. The experiment implements the single-agent exponential bandit model. We analyze how subjects respond to changes in the prior belief, safe action, and discount factor. We find that subjects respond in the predicted direction to these changes. However, we find that subjects under-respond to the prior belief, under-respond to the safe action, and typically explore less than predicted. Our results suggest that neither risk aversion nor the random termination probability are driving under-experimentation. Our results are consistent with subjects having incorrect beliefs about exploration.

AB - This paper analyzes how individuals resolve an exploration versus exploitation trade-off in a laboratory experiment. The experiment implements the single-agent exponential bandit model. We analyze how subjects respond to changes in the prior belief, safe action, and discount factor. We find that subjects respond in the predicted direction to these changes. However, we find that subjects under-respond to the prior belief, under-respond to the safe action, and typically explore less than predicted. Our results suggest that neither risk aversion nor the random termination probability are driving under-experimentation. Our results are consistent with subjects having incorrect beliefs about exploration.

KW - bandits

KW - continuous time

KW - experimentation

KW - exponential bandit model

UR - http://www.scopus.com/inward/record.url?scp=85163028242&partnerID=8YFLogxK

U2 - 10.1111/ecin.13164

DO - 10.1111/ecin.13164

M3 - Article

AN - SCOPUS:85163028242

SN - 0095-2583

VL - 62

SP - 267

EP - 286

JO - Economic Inquiry

JF - Economic Inquiry

IS - 1

ER -

Exploration versus exploitation: A laboratory test of the single-agent exponential bandit model

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Cite this