What Do Applicants Think Employers Want to See in Their Resumes? An Experiment on Second Order Beliefs

How to cite

Zhou, Ying. 2026. "What Do Applicants Think Employers Want to See in Their Resumes? An Experiment on Second Order Beliefs." AEA RCT Registry. AEARCTR-0018294. https://doi.org/10.1257/rct.18294-1.0

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

Ying Zhou*

* Primary investigator

Registry ID
AEARCTR-0018294
DOI
https://doi.org/10.1257/rct.18294-1.0
First registered
Last updated
Status
In development
Trial dates
2026-07-26 – 2029-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/18294

Abstract

Existing research on the gender wage gap has considered gender differences in job attributes as possible explanations. However, the role of workers' belief about what traits employers want from them is underexplored. When seeking jobs, workers are likely to emphasize the achievements that they believe the employer values most from them. Gender differences in these beliefs may explain why women accept different tasks, write resumes differently than men, and ultimately sort themselves into different wages.

Experimental design

I ask experimental subjects to construct short resumes for fictitious male and female workers from a master list of previous achievements, to see which tasks subjects think will yield a job offer. More specifically, subjects need to select and rank achievements (in bullet point form). In the master list, some tasks are promotable (as defined by Babcock et al. 2022), meaning that they generate visible returns to the organization, and some are not. In the second phase of my experiment, I submit the created resumes to real job ads and use the resulting callback rates to reward my subjects. For comparison, I also ask subjects how they think different tasks would contribute to the worker’s job satisfaction.

Design details

Randomization method
Randomization was handled by Qualtrics' built-in features
Randomization unit
The unit of randomization is the individual. Participants will be randomly assigned to see either the male or female version of the same resume content. To see how subjects react to fictitious workers of different genders, the treatment group is defined as those who see the female name, while the control group comprises those who see the male name.
Clusters
1 participant pool
Observations
400 participants
Arms
Although the gender of the subjects is a variable of interest, I cannot pre-determine the exact number of male and female participants. However, randomization should ensure that approximately half of the subjects will see a resume with a female name, while the other half will see a male name.
Primary outcomes
Rank vector of (non-)promotable bullet points that are provided by the subjects in resume design. Outcome variables from full rankings could be (1) whether a promotable bullet point is ranked 1st, 2nd, ..., (2) whether a promotable BP is ranked top1, top2, ..., (3) the average rank of promotable bullet points.
Secondary outcomes
Subjects are also asked to rank how much this fictitious worker enjoyed each of those activities in their current job. The secondary outcome is the subject's ranking of job enjoyment. In the second part of the experiment, resumes will be submitted to real job ads; thus, another secondary outcome is the callback rate.

Cited by

No citation data available for this registration yet.

Keywords