Gender-based Discrimination in Personal Care Services

Last registered on July 20, 2023


Trial Information

General Information

Gender-based Discrimination in Personal Care Services
Initial registration date
July 20, 2023

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
July 20, 2023, 5:57 PM EDT

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



Primary Investigator

University of Massachusetts Amherst

Other Primary Investigator(s)

Additional Trial Information

In development
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
In this study, we explore an under-studied form of discrimination which bases on worker’s visible gender expression and presentation. Gender expression refers to the ways in which workers present or perform their gender identity to their family, relatives, friends, coworkers, and the external social world. Clients' perception of how workers express their gender might interact with occupation-specific gender norms to generate disparity in hiring, retention, and career advancement. This can be especially prevalent in care industries because care is symbolically associated with womanhood and femininity, and care provision is considered “women’s jobs”. In addition, given the prevailing racialized sexual stereotypes in the US, clients' perception of workers' gender expression might also amplify or mediate the effect of racial factors in hiring.

To test for gender expression discrimination, we utilize experimental methods, combining online survey and stimulated decision making. Overall, the study seeks to expand the literature and deepen our understanding of gender-based discrimination, and identity-based discrimination in general. Our experiment highlights the ways in which races, gender, and sexuality overlap and manifest as potential employment barriers, especially for workers in entry-level, lower-paying occupations.
External Link(s)

Registration Citation

Nguyen, Duc Hien. 2023. "Gender-based Discrimination in Personal Care Services." AEA RCT Registry. July 20.
Experimental Details


We create fictional workers' profile with portrait and manipulate the gender expression of the portraits.
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Participants' rating of workers, interviewing rate, and starting wage offer
Primary Outcomes (explanation)
Participants are asked to rate how good a fit the workers are on a scale from 1 to 10
Participants are also asked which workers they want to interview for the job, and conditional on interviewing, the starting hourly wage they would offer to the workers.

Secondary Outcomes

Secondary Outcomes (end points)
Parent's ranking and hiring manager's ranking
Secondary Outcomes (explanation)
Participants are asked to provide a best guess of (1) how their parents would rate the workers, (2) how a hiring manager for a public senior care facility in a "blue" state would rate the worker, (3) how a hiring manager for a religious senior care facility in a "red" state would rate the worker

Experimental Design

Experimental Design
The experiment consists of a fictional hiring scenario for a female-dominated occupation or a gender-neutral occupation. Participants are shown profiles of fictional workers, which include their portraits, age, education attainment, qualifications, experiences, and other details. After viewing the workers, participants will be asked to decide who to hire and the hourly wage rate they want to offer.
Experimental Design Details
The occupation I have selected for this study is personal care attendant (PCA) for the elderly. Like many other care occupations, PCA is often seen as a woman’s job. The ideal PCA is expected to have feminine characteristics such as being kind, warm, emotionally sensitive, compassionate, empathetic, and devoted to taking care of their clients. As such, workers who appear masculine might be perceived as not possessing these desirable qualities and therefore might not be qualified for the job. Unlike the case with nurses or kindergarten teachers where customers are assigned to a service provider, here customers have greater unilateral power to hire or fire the PCA.

The hiring scenario for this position describes a situation where the subjects need to hire a PCA to provide care for their elderly parents. We expect that subjects will display a heightened level of scrutiny because they are making decision on behalf of a more vulnerable population (the elder). The scenario will specify that the main responsibilities of the worker are care-giving tasks, such as providing companionship, cooking, housekeeping, etc. By highlighting these tasks, we want to prime subjects to associate this position with femininity and womanhood, thus increase the likelihood that masculine-presenting workers will be discriminated.

As a control for the experiment, I select group fitness facilitator for senior home facilities. Like PCAs, group fitness instructor provide personal service to their clients and the job requires regular, repeated, in-person interaction. Unlike the PCA, however, group fitness instructors do not have a strong social association with femininity, and thus provide a neutral baseline for comparison (according to recent CPS data, around 85% of PCA are female while only 60% of fitness instructors are female).

To make the hiring scenario for the group fitness instructor comparable to the hiring scenario for the CPA, we will highlight that the subjects need to hire a fitness instructor for a retirement home where their parents reside. The fitness classes are meant to keep seniors active, mobile, and offer opportunities for socialization. We pick these dimensions to prime the subjects to think of this job as more gender neutral, as opposed to being overly physical and masculine, like a personal body-building trainer.
Randomization Method
Randomization by built-in random number generator on Qualtric
Randomization Unit
- Participants are randomly assigned to the control experimental session or treatment experimental session.
- Workers' details (other than their portrait) are also randomly assigned at the participant level.
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
1,000 - 1,500 participants
Sample size: planned number of observations
4,000 - 6,000 data point at the worker - participant level.
Sample size (or number of clusters) by treatment arms
500 - 750 participants will be assigned to the treatment group, generating between 2,000- 3,000 data point
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
University of Massachusetts Amherst
IRB Approval Date
IRB Approval Number


Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information


Is the intervention completed?
Data Collection Complete
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials