To what extent does the choice to acquire information determine learning?

How to cite

Grabowska, Marta; Lagarde, Mylene. 2023. "To what extent does the choice to acquire information determine learning?." AEA RCT Registry. AEARCTR-0012401. https://doi.org/10.1257/rct.12401-3.0

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

Marta Grabowska* , Mylene Lagarde

* Primary investigator

Registry ID
AEARCTR-0012401
DOI
https://doi.org/10.1257/rct.12401-3.0
First registered
Last updated
Status
Completed
Trial dates
2024-01-31 – 2024-08-30
Countries
Ghana ()
Registry record
http://www.socialscienceregistry.org/trials/12401

Abstract

This study explores the impact of (no) choice in information acquisition on learning outcomes in the context of health communication. Decisions about investments into health are often made under uncertainty by individuals who lack perfect information. Information provision interventions are used to counteract this by supplying accurate information. However, these interventions often fail to improve participants’ knowledge or shift their decisions towards the optimum. This study investigates one potential reason for this: the fact that the way that information is acquired affects the extent of learning. I study whether giving individuals a choice in accessing information about the new malaria vaccine affects their attention to the information and their resulting knowledge about, and support for, the vaccine. I do this in a survey experiment conducted in Ghana where I randomly assign participants to 1) passively watch a placebo video about child development; 2) passively watch the treatment video about the malaria vaccine; 3) the choice group, where participants express a preference between the placebo and treatment videos. In group 3, there is a further randomisation of participants following the choice; some watch the video they preferred, while others watch the video they did not prefer. I compare how individuals respond to the treatment video if they watch it passively, watch it "keenly" (after choosing it), or watch it "forcedly" (after choosing against it). Through this study, I seek to inform health communication strategies, particularly in the context of emerging health interventions.

Intervention

There are two information videos used in the experiment: the ``malaria'' video, which provides information about malaria in children and the new vaccine, and the ``IDS'' or infant-directed speech video, which is an active control video\footnote{The videos in different languages were developed and made publicly available by \citet{dupas2023informing}, to whom we are deeply grateful for their permission to re-use the content in this study.}. The ``malaria'' video covers information such as malaria mortality and morbidity figures, including the burden in Ghana; effects of malaria on the community; the benefits of the vaccine; and other malaria prevention strategies. The video is an animation voiced over in several languages by local actors to ensure each participant receives the content in the vernacular. The control ``IDS'' video is very similar in format, design style and duration, and is likewise translated for each region. The only difference is in content: the video explains that talking to babies from a young age helps them develop cognitively, provides practical examples of how to fit IDS into your life, and how to engage others in the household in talking to the baby.

Experimental design

Participants who consent to take part in the study will be randomly assigned to one of 3 intervention arms:

1. Placebo arm: participants passively (i.e., with no choice of video offered) watch a placebo information video about an unrelated topic on child care. The video is very similar in design to the malaria video but delivers none of the information. This arm consists of 560 participants.
2. Passive treatment: participants passively watch the intervention video. This arm consists of 564 participants.
3. Offered a choice: participants are asked for their preference between the placebo and the intervention videos. They learn that their choice is more likely to be shown, though it is still randomised by the survey device. There are 1,654 participants in this arm.

A second randomisation then takes place, assigning participants to either their choice video or not. This creates four further intervention arms:
3.1 Those who wanted to view the intervention video and are shown it ("chosen" viewers) - 558 participants
3.2 Those who wanted to view the placebo video but are shown the intervention video ("forced" viewers) - 824 participants
3.3 Those who wanted to view the placebo video and are shown it - excluded from analysis, 203 participants
3.4 Those who wanted to view the intervention video but are shown the placebo video - excluded from analysis, 69 participants

Design details

Randomization method
Randomisation done within each survey device according to pre-specified randomisation proportions. We use SurveyCTO.
Randomization unit
Individual
Clusters
0
Observations
2,506 individuals
Arms
1. Placebo arm: 560. 2. Passive watch: 564. 3. Chosen watch: 558. 4. Forced watch: 824.
Primary outcomes
1. Attention to information measured through a summary index of video recall and knowledge recall. 2. Malaria vaccine efficacy beliefs measured as a likelihood out of 10. 3. Intention to vaccinate own children, if eligible in the future.
Secondary outcomes
1. Endorsement of the malaria vaccine 2. Demand for further information about the vaccine
Sponsors
Name: The Mercury Project, Social Science Research Council Location: USA Url: https://www.ssrc.org/programs/the-mercury-project/
Partners
Name: Innovations for Poverty Action (Ghana) Partner_type: ngo Url: https://poverty-action.org/ghana-overview

Cited by

No citation data available for this registration yet.

Keywords