Effectiveness of a Mayan language social media campaign on COVID-19 vaccine beliefs in Guatemala

How to cite

Johnston, Jamie. 2022. "Effectiveness of a Mayan language social media campaign on COVID-19 vaccine beliefs in Guatemala." AEA RCT Registry. AEARCTR-0008993. https://doi.org/10.1257/rct.8993-1.0

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

Jamie Johnston*

* Primary investigator

Registry ID
AEARCTR-0008993
DOI
https://doi.org/10.1257/rct.8993-1.0
First registered
Last updated
Status
In development
Trial dates
2022-03-01 – 2022-04-01
Countries
Guatemala ()
Registry record
http://www.socialscienceregistry.org/trials/8993

Abstract

Prior evidence, including formative work supporting the development of this study, suggests there is widespread vaccine hesitancy among Indigenous communities in Guatemala, fueled by mistrust in the health system, lack of official information, and the circulation of myths and misinformation. We will test the effectiveness of a video-based social media campaign that provides a basic overview of the science behind COVID-19 vaccines and addresses prevalent myths and misinformation being shared in target communities. A series of animated videos will be promoted through Facebook Ads and randomized at the individual level across Facebook users throughout the entire country. To investigate the effectiveness of Mayan language content, we are testing three treatment arms – visually identical videos in the 1) Spanish language, 2) K'iche' language, and 3) Kaqchikel language. Our primary outcomes are responses to two attitudinal questions collected via Facebook polling: 1) How safe do you think a COVID-19 vaccine is for people like you? (options: very safe, somewhat safe, barely safe, not safe, don’t know); and 2) When you think of most people whose opinion you value, how much would they approve of people getting a COVID-19 vaccine? (options: definitely approve, mostly approve, somewhat approve, not at all approve, don’t know). We will measure effects of the three treatment arms across Spanish speakers (including K’iche’ and Kaqchikel speakers who also speak Spanish). We will also measure the effectiveness of K’iche vs. Spanish content among K’iche speakers and Kaqchikel vs. Spanish content for Kaqchikel speakers.

Intervention

A series of animated video clips will be promoted through Facebook Ads and randomized at the individual level across Facebook users throughout the entire country. To investigate the effectiveness of Mayan language content, we are testing three treatment arms – visually identical videos in the 1) Spanish, 2) K'iche', and 3) Kaqchikel languages. The videos are visually identical and follow the same audio script, translated into the three focal language. Local voiceover artists were used to record the scripts in the three languages.

Experimental design

Facebook Users are randomized at the individual level, selected by Facebook's algorithms to be exposed to the ad campaign view videos on their Facebook feeds approximately X times during the duration of the campaign. Individuals in the control groups will not be exposed to any ad campaign videos.

Design details

Randomization method
Computerized randomization is being done at the individual level via Facebook's Brand Lift tool.
Randomization unit
individual
Clusters
N/A
Observations
The Facebook Brand Lift tool will randomize across all Facebook users in Guatemala. We approximate that 6 million Facebook users in Guatemala will be exposed to the ads.
Arms
Randomization across treatment arms will be roughly equal -- among Spanish-speaking treatment, roughly 1/3 will receive Spanish language content, 1/3 will receive K'iche' language content, and 1/3 will receive Kaqchikel language content. Among K'iche- / Kaqchikel speaking treatment, 1/2 will receive Spanish language content and 1/2 will K'iche- / Kaqchikel language content, respectively.
Primary outcomes
Our primary outcomes are responses to two attitudinal questions collected via Facebook polling: 1) How safe do you think a COVID-19 vaccine is for people like you? (options: very safe, somewhat safe, barely safe, not safe, don’t know); and 2) When you think of most people whose opinion you value, how much would they approve of people getting a COVID-19 vaccine? (options: definitely approve, mostly approve, somewhat approve, not at all approve, don’t know).

Cited by

No citation data available for this registration yet.

Keywords