Individual versus group work

How to cite

Coutts, Alexander; Dinc, Sukran. 2025. "Individual versus group work." AEA RCT Registry. AEARCTR-0015798. https://doi.org/10.1257/rct.15798-2.0

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

Alexander Coutts* , Sukran Dinc

* Primary investigator

Registry ID
AEARCTR-0015798
DOI
https://doi.org/10.1257/rct.15798-2.0
First registered
Last updated
Status
In development
Trial dates
2025-04-23 – 2026-05-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/15798

Abstract

Collaboration is essential in education and the workplace but introduces challenges related to uncertainty and unequal recognition, particularly in evaluating individual contributions. Preferences for working collaboratively or independently may vary by gender, performance, and perceptions of bias, with evidence suggesting women often receive less credit in team settings. This experiment investigates both individuals’ preferences for collaboration and evaluators’ decisions in assessing performance. Participants choose between working alone or in teams, while evaluators select among candidates with different work modes. Two experimental conditions manipulate the visibility of gender to assess its impact on evaluation outcomes.

Intervention

The study investigates (i) individuals’ preferences to work in a pair versus individually and (ii) third-party evaluations of individual performance based on observed output. In the gender treatment, evaluators hear voices that reveal the gender of participants; in the baseline, voices are gender-neutral. Full details are described in the attached pre-analysis plan.

Experimental design

This is a two-part, incentivized online experiment. In Experiment 1, participants (“workers”) decide whether to work individually or in a pair on a Raven’s test, and how much influence to exert over the pair's score if paired. In Experiment 2, another group of participants (“evaluators”) chooses among three past workers (1 solo, 2 paired) based on observable scores, certificates, and voice cues. Treatments vary whether gender is revealed and whether the assignment to individual/pair work was endogenous (participant-selected) or exogenous (random). Full experimental details and protocols are included in the pre-analysis plan.

Design details

Randomization method
Randomization is conducted by computer using Qualtrics' built-in randomization logic, based on pre-assigned quotas and random number generators.
Randomization unit
The unit of randomization is the individual participant. Workers and evaluators are randomized into treatment conditions independently. Matching occurs at the individual level but constrained by pre-specified treatment cells (e.g., gender visibility, certificate type, endogenous/exogenous assignment).
Clusters
Not applicable (individual-level randomization only).
Observations
2,800 total participants: 1,600 workers (800 in gender treatment, 800 in baseline). 1,200 evaluators (800 in gender treatment, 400 in baseline). Sample size by may vary slightly due to random assignment procedure.
Arms
Workers: 800 in gender treatment (400 men, 400 women); 800 in baseline (400 men, 400 women). Evaluators: 800 in gender treatment (400 men, 400 women); 400 in baseline (200 men, 200 women). All cells are further cross-randomized with endogenous/exogenous assignment (approx. 50/50 split). Sample size by treatment may vary slightly due to random assignment procedure.
Primary outcomes
Workers' choice to work individually vs. in a pair. Workers’ willingness to pay to override an assigned (non-preferred) work arrangement. Workers’ contribution weight when forming a pair score. Evaluators' selection: whether they choose an individual worker or a worker from a pair. Evaluators' selection between paired workers (based on certificate type and gender). See the pre-analysis plan for operationalizations and measurement procedures.

Cited by

No citation data available for this registration yet.

Keywords