How Is Your User Feeling? Inferring Emotion Through Human-Computer Interaction Devices

MIS Quarterly, 41(1): pp. 1-21, 2017

48 Pages Posted: 4 Jan 2016 Last revised: 18 Mar 2018

Martin Thomas Hibbeln

University of Duisburg-Essen - Mercator School of Management

Jeffrey L. Jenkins

University of Arizona

Christoph Schneider

City University of Hong Kong (CityUHK)

Joseph Valacich

University of Arizona

Markus Weinmann

University of Liechtenstein

Date Written: 2017

Abstract

Experiencing negative emotion during system use can adversely influence important user behaviors, including purchasing decisions, technology use, and customer loyalty. The ability to easily assess users' negative emotion during live system use therefore has practical significance for the design and improvement of information systems. We utilize attentional control theory to explain how mouse cursor movements can be a real-time indicator of negative emotion. We report three studies. In Study 1, an experiment with 65 participants from Amazon's Mechanical Turk, we randomly manipulated negative emotion and then monitored participants' mouse cursor movements as they completed a number-ordering task. We found that negative emotion increases the distance and reduces the speed of mouse cursor movements during the task. In Study 2, an experiment with 126 participants from a U.S. university, we randomly manipulated negative emotion and then monitored participants' mouse cursor movements while they interacted with a mock e-commerce site. We found that mouse cursor distance and speed can be used to infer the presence of negative emotion with an overall accuracy rate of 81.7%. In Study 3, an observational study with 80 participants from universities in Germany and Hong Kong, we monitored mouse cursor movements while participants interacted with an online product configurator. Participants reported their level of emotion after each step in the configuration process. We found that mouse cursor distance and speed can be used to infer the level of negative emotion with an out-of-sample R2 of 0.17. The results enable researchers to assess negative emotional reactions during live system use, examine emotional reactions with more temporal precision, conduct multi-method emotion research, and create more unobtrusive affective and adaptive systems.

Keywords: negative emotion, attentional control theory (ACT), mouse cursor distance, mouse cursor speed, mouse tracking, human-computer interaction

JEL Classification: L86

Suggested Citation

Hibbeln, Martin Thomas and Jenkins, Jeffrey L. and Schneider, Christoph and Valacich, Joseph and Weinmann, Markus, How Is Your User Feeling? Inferring Emotion Through Human-Computer Interaction Devices (2017). MIS Quarterly, 41(1): pp. 1-21, 2017. Available at SSRN: https://ssrn.com/abstract=2708108

Martin Thomas Hibbeln

University of Duisburg-Essen - Mercator School of Management ( email )

Lotharstra├če 65
Duisburg, Nordrhein-Westfalen 47057
Germany
+49 203 379-2830 (Phone)

Jeffrey L. Jenkins (Contact Author)

University of Arizona ( email )

Department of History
Tucson, AZ 85721
United States

Christoph Schneider

City University of Hong Kong (CityUHK) ( email )

83 Tat Chee Avenue
Kowloon
Hong Kong

Joseph Valacich

University of Arizona ( email )

Department of History
Tucson, AZ 85721
United States

Markus Weinmann

University of Liechtenstein ( email )

F├╝rst-Franz-Josef-Strasse
Vaduz, 9490
Liechtenstein

Register to save articles to
your library

Register

Paper statistics

Downloads
315
rank
86,646
Abstract Views
1,099
PlumX