Using nudges to improve participation in worker well-being events. A field study.

How to cite

Pelly, Diane; Doyle, Orla. 2021. "Using nudges to improve participation in worker well-being events. A field study.." AEA RCT Registry. AEARCTR-0008479. https://doi.org/10.1257/rct.8479-1.1

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

Diane Pelly* , Orla Doyle

* Primary investigator

Registry ID
AEARCTR-0008479
DOI
https://doi.org/10.1257/rct.8479-1.1
First registered
Last updated
Status
Completed
Trial dates
2021-11-04 – 2022-03-31
Countries
Ireland ()
JEL codes
C93; D11; M54; J01
Registry record
http://www.socialscienceregistry.org/trials/8479

Abstract

Organisations are increasingly investing substantial time and money in the design and implementation of initiatives aimed at boosting the emotional and psychological well-being of their employees. While well-intentioned, the impact of such initiatives depends on the extent to which their employees engage with and use them. The aim of this trial is to test whether ‘nudges’ (Thaler & Sunstein, 2008) which have been shown to be effective at improving attendance in other contexts (e.g., Fishbane, Ouss & Shah, 2020; Dai et al. 2020) can be used to extend the reach of well-being initiatives within an organisational context, thus improving return on investment. We employ a RCT design to test the impact of four behavioural science nudge techniques (simplification, social proof, harnessing the default and changing the messenger), on the registration and participation rates of 7,000 employees in relation to three online ‘Lunch ‘n Learn’ well-being events being organised by a large Irish public sector organisation.

Intervention

We will be testing the extent to which four standard nudge techniques (simplification; changing the messenger; harnessing defaults and social proof) are effective at boosting registration and attendance rates relating to three online well-being events being organised by a large Irish public sector organisation. The control group will be issued with the standard email invitation to attend the event. The four treatment groups will be issued modified versions of the standard email invitation. Treatment Group 1 (Simplification) will receive an email invitation which contains an embedded direct link to registration. Treatment Group 2 (Messenger) will receive an email invitation signed by a senior manager in the organisation who endorses the events and advocates attendance. Treatment Group 3 (Default) will receive an email informing them that they have been selected to attend one of the 3 well-being events and giving them the option to either 1) do nothing and receive the link to the online event (i.e. opt-in); or 2) click on a link in the email which enables them to either switch to one of the other 3 events or indicate that they do not wish to attend any event (i.e. opt-out). Treatment Group 4 (social proof) will receive an email reminder a few days prior to the events, which informs them that X% of their colleagues have already signed up and urges them to do likewise.

Experimental design

A RCT design will be employed. There will be five randomly assigned (equal proportions) conditions – the control and four treatment arms. All 7,000 employees in the organisation will be invited by email to attend three diversity and inclusion themed ‘Lunch ‘n Learn’ online events which are being organised by the organisation’s well-being committee. All employees will be sent one follow-up reminder to sign-up for the event 1-4 days prior to the event. The control group will receive the standard invitation email from the well-being committee to attend the events, which provides some information on the events and asks people who are interested in attending to confirm by return email that they would like to be added to the attendance list and therefore be sent a follow-up electronic link to enable them to attend the online event. Treatment group 1 (simplification) will receive an invitation email from the well-being committee which enables them to register for the event(s) using an online link which is embedded in the email. Treatment group 2 (messenger) will receive the same email as Treatment group 1 but with the addition of a signed message from a senior manager in the organisation which endorses the events, confirms that he/she will be getting involved, and urges other employees in Treatment group 2 to do the same. Treatment group 3 (default) will receive the same email as Treatment group 1 but with the additional of some extra text which informs the recipient that they have been randomly selected to attend one of the three events and gives them the option to either click on a link which will enable them to either switch to one of the other two events or to opt-out completely or to do nothing, in which case they will be emailed a link to the online event at a later date. Treatment group 4 (social proof) will receive the same email invitation as treatment group 1 but will also receive a second follow-up email before the event informing them that x% of their colleagues have signed-up for one or more of the events and urging them to do likewise.

Design details

Randomization method
Participants will be randomly signed (using an individual probability randomisation strategy i.e. simply randomisation with no blocking or stratification) into one of 5 groups using STATA software. In addition, Treatment group 3 (default) will be randomly opted in to attend one of the three scheduled events.
Randomization unit
Individual
Clusters
Zero
Observations
7,000
Arms
Control: 1,400 employees Treatment group 1 (Simplification): 1, 400 employees Treatment group 2 (Messenger): 1,400 employees Treatment group 3 (Default): 1,400 employees Treatment group 4 (Social Proof): 1,400 employees
Primary outcomes
o % of workers who register to attend at least one of three scheduled online well-being events o % of workers who actually attend at least one of three scheduled online well-being events

Cited by

No citation data available for this registration yet.

Keywords