Replacing Humans by Robots: An Experiment on Perceived Trustworthiness

Last registered on May 21, 2024

View Trial History

Pre-Trial

Trial Information

General Information

Title

Replacing Humans by Robots: An Experiment on Perceived Reliability

RCT ID

AEARCTR-0013546

Initial registration date

May 17, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

May 21, 2024, 11:33 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Switzerland

Region

Zurich

Primary Investigator

Name

Carlos Gomez Gonzalez

Affiliation

University of Zurich

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Yu Pan

PI Affiliation

University of Zurich

Contact Investigator

PI Name

Helmut Dietl

PI Affiliation

University of Zurich

Contact Investigator

PI Name

Gwen-Jiro Clochard

PI Affiliation

Osaka University

Contact Investigator

Additional Trial Information

Status

In development

Start date

2024-05-28

End date

2024-06-04

Keywords

Behavior, Labor, Other

Additional Keywords

JEL code(s)

C91; D91; L80

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

We investigate how robots influence perceived reliability in various scenarios, including hotels, restaurants, and dental clinics. Our study examines the impact of mechanical and thinking/feeling tasks performed by robots on customer perceived reliability. Additionally, we assess whether enhancements in a robot's physical appearance, exemplified by the Tesla robot, and the presence of human supervision can improve perceptions of tasks rendered by robots. In this analysis plan, we pre-register the experimental framework we will follow.

External Link(s)

Registration Citation

Citation

Clochard, Gwen-Jiro et al. 2024. "Replacing Humans by Robots: An Experiment on Perceived Reliability." AEA RCT Registry. May 21. https://doi.org/10.1257/rct.13546-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

We perform three experiments with Prolific Online workers to examine how customers' perceived reliability is influenced by robots. Additionally, we test whether enhancements in physical appearance and human supervision moderate the effect.

Intervention (Hidden)

Intervention Start Date

2024-05-28

Intervention End Date

2024-06-04

Primary Outcomes

Primary Outcomes (end points)

The main variable of interest is the overall perceived reliability.

Primary Outcomes (explanation)

We measure overall satisfaction using a 7-point Likert scale. Participants strongly agree/disagree with the following statement:
"Overall, I find restaurant/hotel/dental clinic to be reliable".

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

In Study 1, participants view daily life videos depicting three scenarios, i.e., dental clinic, restaurant, and hotel, where humans perform mechanical and feeling/thinking tasks. We investigate participant perceptions of reliability when robots replace humans performing these tasks. We examine three robot conditions (between-subjects): Mechanical robots (Robots perform mechanical tasks) vs. Feeling/thinking robots (robots perform feeling/thinking tasks) vs. All-Robots (Robots perform both tasks)

In Study 2, we replace the Pepper robot from Study 1 with a more technologically advanced Tesla robot to examine if superior aesthetics enhance perceived reliability.

In Study 3, we extend the scenario description from Study 1, informing participants that robots are working under staff supervision. This information element assesses the impact of human supervision on perceived reliability.

Experimental Design Details

Randomization Method

Our experimental framework employs a 3 (robot condition, between-subjects: Mechanical Robots vs. Thinking/Feeling Robots vs. All-Robots) by 3 (scenario video, within-subjects: Hotel vs. Restaurant vs. Dental Clinic) mixed design across three studies. Participants are randomly assigned to Study 1, 2, or 3, and then randomly assigned to one of the three robot conditions, in which they are presented with three scenario videos.

Randomization Unit

Participant level.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Our study does not involve clusters; participants are individually assigned to different treatment conditions.

Sample size: planned number of observations

2000 participants

Sample size (or number of clusters) by treatment arms

800 participants in Study 1 (Pepper Robot, No Human Supervision)
600 participants in Study 2 (Tesla Robot, No Human Supervision)
600 participants in Study 3 (Pepper Robot, With Human Supervision)

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Study 1 enables the detection of an effect size of 0.017 with 80% power at an alpha level of 0.01 in a linear regression with 9 predictors. Study 2 enables the detection of an effect size of 0.017 with 80% power at an alpha level of 0.01 in a linear regression with 13 predictors. Study 3 enables the detection of an effect size of 0.017 with 80% power at an alpha level of 0.01 in a linear regression with 13 predictors.

Supporting Documents and Materials

IRB