Statistics and Stories: Shaping Memory

How to cite

Adam, Salamatu Nanna. 2024. "Statistics and Stories: Shaping Memory." AEA RCT Registry. AEARCTR-0013880. https://doi.org/10.1257/rct.13880-1.0

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

Salamatu Nanna Adam*

* Primary investigator

Registry ID
AEARCTR-0013880
DOI
https://doi.org/10.1257/rct.13880-1.0
First registered
Last updated
Status
In development
Trial dates
2024-07-05 – 2024-08-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/13880

Abstract

We propose to study how risk information about HIV should be presented in order to maximize its effect on beliefs about HIV for longer periods. In our baseline experiment, we will provide a random sample of participants with quantitative (statistical) risk information about the national new HIV infections and the attendance rates at a local clinic for HIV testing. For the remaining participants, we will complement this statistical information with a story, that is a qualitative narrative explaining how someone overcame the fear and anxiety to get tested after an ex-partner tested positive for HIV. The story also informs the reader about the testing experience at a local clinic. We will give all participants vouchers for HIV tests. This experimental design will allow us to estimate the effect of stories on the evolution of beliefs about HIV. Next, we will estimate the effect of stories on recall of the statistical information received two weeks after information is delivered. Lastly, We will estimate the effect of stories on voucher take-up intentions and actual HIV tests conducted at the clinic.

Experimental design

We propose to conduct a field experiment in Ghana with a student sample to study how the presentation of risk information about HIV affects long-term beliefs about the virus. The experiment will consist of two parts.

In the first part, we will provide a random sample of participants with quantitative (statistical) risk information about the number of new HIV infections recorded in Ghana and the attendance rates at a local clinic for HIV testing. For some participants, we will complement this statistical information with a qualitative narrative. This narrative will describe how someone overcame fear and anxiety to get tested after an ex-partner tested positive for HIV and will include details about the testing experience at the clinic. We will vary the narrator's test results (positive or negative) to examine which types of stories are most effective.

Before providing the information treatment, we will record participants' demographic data and elicit their prior beliefs about the number of new HIV infections in Ghana and the percentage of HIV tests conducted at the clinic. After the information treatment, we will measure their posterior beliefs about these metrics again. Additionally, we will offer HIV test vouchers and document participants' intentions to use them.

Two weeks after the first part of the experiment, we will implement the second part. In this phase, we will measure participants' posterior beliefs about the number of new HIV infections in Ghana and the percentage of HIV tests conducted at the clinic again. We will also conduct an incentivized guess and recall task based on the information provided in the first part of the survey. Finally, we will measure HIV test voucher take-up intentions again and distribute HIV test vouchers to all participants.

Design details

Randomization method
Randomization done by Qualtrics
Randomization unit
Unit of randomization is individual
Clusters
2400 individuals
Observations
2400 individuals
Arms
600 individuals in control, 600 individuals in positive story treatment (shown story of someone who tested positive for HIV as well as national new HIV infection rates), 600 individuals in negative story treatment (shown story of someone who tested negative for HIV as well as national new infection rates), 600 individuals in statistics treatment (shown information about the national new infection rates)
Primary outcomes
Our primary outcomes of interest are (i) incentized posterior beliefs about the percentage of HIV tests conducted at the clinic in the year 2023 recorded both immediately after the main survey and the follow up and (ii) respondents decision to take an HIV test after the survey. After the follow up, respondents will receive a voucher which they can exchange for a free HIV screening at a clinic.
Secondary outcomes
Our secondary outcomes are HIV test voucher take-up intentions measured both immediately after the main survey and at the follow, and recall rates of information about the 2022 national new HIV infections.

Cited by

No citation data available for this registration yet.

Keywords