AI as an Employee: How AI Affects Managers’ Evaluation of Creative Works

How to cite

Meng, Juanjuan. 2024. "AI as an Employee: How AI Affects Managers’ Evaluation of Creative Works." AEA RCT Registry. AEARCTR-0014071. https://doi.org/10.1257/rct.14071-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Juanjuan Meng*

* Primary investigator

Registry ID
AEARCTR-0014071
DOI
https://doi.org/10.1257/rct.14071-1.0
First registered
Last updated
Status
In development
Trial dates
2024-07-29 – 2024-08-08
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/14071

Abstract

This study aims to explore how the introduction of AI as an employee within an organization will affect managers' evaluation of other employees and their ultimate decision-making. More specifically, our research scenario involves scoring creative works. We plan to have two experimental groups. In Group1, only the initial evaluation scores, i.e., the average score given during the preliminary evaluation by subordinate employees are provided to managers. In Group2, both AI scores and the initial evaluation scores are provided. We will observe the final score given by managers and their decision time. The scoring involves three dimensions: target and difficulty; creativity and execution; and outcome. We will analyze the heterogeneity impacts of providing AI score by the above dimensions. This study will provide implications on how managers (experts) react to AI, with a focus on creativity industry.

Intervention

Experimental Group 1: Only the initial evaluation scores are provided (i.e., the average score given during the preliminary evaluation by subordinate employees).
Experimental Group 2: Both AI scores and the initial evaluation scores are provided.

Experimental design

1. Scenario:
Judges log in to the scoring system to rate creative works, with each judge scoring between 20 to 50 pieces.

2. Groups:
Experimental Group 1: Only the initial evaluation scores are provided (i.e., the average score given during the preliminary evaluation by subordinate employees).
Experimental Group 2: Both AI scores and the initial evaluation scores are provided.

3. Participants:
The judges are approximately 90 senior executives from the marketing industry.

4. Experimental Procedure
Develop the scoring system to display reference scores according to the experimental design.
Once the judges begin scoring, record the scores and the time taken to score.

Design details

Randomization method
randomization done in office by a computer
Randomization unit
individual
Clusters
95 individuals
Observations
95 individuals
Arms
about 45 individuals per treatment arm
Primary outcomes
the score given to the creative works by managers; the decision time.

Cited by

No citation data available for this registration yet.

Keywords