Misconduct reporting and identity

How to cite

Chen, Yutong; Vogelsang, Timo; Ramasubramanian, Hari. 2025. "Misconduct reporting and identity." AEA RCT Registry. AEARCTR-0015610. https://doi.org/10.1257/rct.15610-1.0

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

Yutong Chen* , Timo Vogelsang , Hari Ramasubramanian

* Primary investigator

Registry ID
AEARCTR-0015610
DOI
https://doi.org/10.1257/rct.15610-1.0
First registered
Last updated
Status
Completed
Trial dates
2025-04-17 – 2025-04-18
Countries
Germany ()
Registry record
http://www.socialscienceregistry.org/trials/15610

Abstract

This study examines how individual and organizational identities influence employees’ decisions to report misconduct. Using an online experiment with approximately 2400 participants, it manipulates identity through two key mechanisms. First, team identity, which is established via the Klee and Kandinsky task, in which participants have to choose a painting. Second, organizational identity is established by a financial donation to either the Red Cross or the National Rifle Association (NRA). Managers complete a problem-solving task and report their scores, while employees decide whether to report potential misreporting of these scores. The study explores whether organizational identity can mitigate the negative effect of reporting misreporting that arises from strong team identities. By analyzing how identity-based factors shape behavior to report misreporting, the research provides insights into using identity as a cultural control mechanism.

Experimental design

1. Participant Recruitment and Role Assignment
We recruit UK participants on Prolific. Participants are randomly assigned to one of two roles, Manager or Employee.

2. Work Stage I: Preference Task
Participants complete an individual preference task by choosing their preferred painting from works by Klee and Kandinsky. The Manager’s choice is used to represent the department.

3. Organizational Assignment and Donation Information
Each Manager–Employee pair is later matched to an organization that donates either to the NRA or the Red Cross, determined randomly. Only Employees receive this donation information.

4. Work Stage II: Manager Cognitive Task
Managers independently complete Raven Progressive Matrices questions. Their actual performance is recorded.

5. Work Stage III: Reporting Stage
Managers then state how many questions they believe they answered correctly. Then, Misreported managers will be matched with Employees. Later, Employees are shown both the Manager’s stated score and their actual score and must decide whether to report the Manager for misconduct.

6. Work Stage IV: Post-Experimental Questionnaire
All participants complete a post-experiment questionnaire.

Design details

Randomization method
This is an online experiment. Participants are supposed to come from UK and will be hired via Prolific and they'll randomly form pairs.
Randomization unit
individual
Clusters
The study involves 2400 participants, with 2000 managers and 400 employees. From out pilot experiment, we expect that 20% of Managers will misreport, and the 400 misreporting Managers will form random pairs with the 400 Employees, with each pair independently randomized to experimental conditions in a 2×2 factorial design (match/no match with manager × match/no match with organization). The experimental units are these individual employee-manager pairs.
Observations
400 pairs. Note that the figure of 400 misreporting managers is an estimate determined by the assumed misreporting rate of 20% (in our pilot test, the misreporting rate is 22%); the actual number of misreporting managers, and hence the number of pairs assigned to the experiment, may differ from this projection.
Arms
2400 total participants (around 400 valid employee-manager pairs) The design is a 2×2 factorial design, so there are 4 treatment conditions: (1) Match with Manager + Match with Organization: Approximately 100 employee-manager pairs (2) Match with Manager + Not Match with Organization: Approximately 100 employee-manager pairs (3) Not Match with Manager + Match with Organization: Approximately 100 employee-manager pairs (4) Not Match with Manager + Not Match with Organization: Approximately 100 employee-manager pairs
Primary outcomes
whistleblowing likelihood

Cited by

No citation data available for this registration yet.

Keywords