Rejection timing in the job market and candidate satisfaction

How to cite

Huang, Lidingrong; Smeets, Paul. 2023. "Rejection timing in the job market and candidate satisfaction." AEA RCT Registry. AEARCTR-0011341. https://doi.org/10.1257/rct.11341-2.0

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

Lidingrong Huang* , Paul Smeets

* Primary investigator

Registry ID
AEARCTR-0011341
DOI
https://doi.org/10.1257/rct.11341-2.0
First registered
Last updated
Status
On going
Trial dates
2023-04-28 – 2023-09-27
Countries
Philippines ()
JEL codes
C93, D91, J20
Registry record
http://www.socialscienceregistry.org/trials/11341

Abstract

Preparing job applications is time consuming. Applicants expect their application files to be reviewed carefully and fairly for the effort they put in. A fast rejection soon after submitting the application understandably generates frustration and leads to the perception of being unfairly treated, negative reciprocity follows. After making the determination that a certain candidate is unsuitable, the timing as to when the rejection letter should be sent is then of importance for a good company image, positive candidate experience and job-seekers' morale.

Intervention

Job applicants constantly apply to this nationally leading firm in the debt recovery industry. Applicants are first assessed by computer algorithm based on the online tests and scores before being passed onto (human) HR staffers. Most of the applicants would be automatically rejected for being objectively 'unqualified' for the role.

Randomisation is done at the individual applicant level based on their unique candidate IDs, the only intervention is to vary the timing of sending the generic rejection letter.

The control variables will be extracted from their application packages and the dependent variables are obtained from their responses to a survey contained within the rejection letter. Their decision to whether complete the survey by exerting some effort would also be used as a binary outcome variable.

Experimental design

The subjects are unqualified job applicants that have to be rejected. The rejection letters are generic and the only experimental manipulation here is the timing as to when we send the rejection letters.

Design details

Randomization method
Randomisation at the individual applicant level using a compute algorithm in the office.
Randomization unit
Individual applicant
Clusters
10,000
Observations
10,000
Arms
5,000
Primary outcomes
Only a small number of rejected applicants would fill the survey but it serves as the only way for them to share their dissatisfactions and complaints directly with the firm - a way of emotion venting. We will keep the experiment running until a sufficient amount of sample is received. Questions that included in the survey are conventional candidate application experience questions with Likert scales: 1. How satisfied are you with the application experience? 2. How likely would you encourage your peers and friends to apply to us? 3. We hope to be in touch again when more suitable roles appear, how likely would you accept our invitation to reapply in the future? 4. Open question on what we can do better to improve the recruitment process.
Secondary outcomes
Binary variable on whether applicants decide to answer the survey.

Cited by

No citation data available for this registration yet.

Keywords