To ask or to observe: the impact of baseline measurement on future sanitation behavior

How to cite

Treurniet, Mark; Langley, Madeline; Saxena, Kritika. 2024. "To ask or to observe: the impact of baseline measurement on future sanitation behavior." AEA RCT Registry. AEARCTR-0013686. https://doi.org/10.1257/rct.13686-1.0

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

Mark Treurniet* , Madeline Langley , Kritika Saxena

* Primary investigator

Registry ID
AEARCTR-0013686
DOI
https://doi.org/10.1257/rct.13686-1.0
First registered
Last updated
Status
In development
Trial dates
2024-05-25 – 2027-12-31
Countries
Bangladesh ()
Registry record
http://www.socialscienceregistry.org/trials/13686

Abstract

Baseline data collection is a common practice among empirical researchers. While asking questions is often cheaper, observational data is typically trusted more. In this study we develop an experiment in Bangladesh where we randomly assign respondents to (i) only self-reported questions on sanitation behavior, (ii) a combination of observation and self-reported questions on sanitation behavior, or (iii) no questions on sanitation behavior. Our first objective is to determine if observational data influences self-reported responses at baseline. Subsequently, we will examine whether self-reported questions and observational data impact both self-reported and observed sanitation measures at the endline. Additionally, we will investigate whether self-reported questions and observation affect the impact of a separate sanitation intervention.

Intervention

In the self-report arm, we will ask sanitation questions related to waste disposal (from latrine, diapers and garbage) and handwashing. In the observation arm, before we ask the same sanitation questions, we will first ask whether we can observe, respectively, the respondents’ waste disposal facilities and handwash area.

Experimental design

We have 3 baseline survey arms:
● Group 1: [R] only reported questions
● Group 2: [R+O] both reported and observation questions
● Group 3: [C] Pure control group that gets not questions on sanitation and health in the baseline

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Individual randomization, stratified by the group-level treatment assignment of the larger RCT within which this survey experiment takes place
Clusters
1034 individuals
Observations
1034 individuals
Arms
343 individuals in group 1, 344 in group 2, 347 in group 3
Primary outcomes
Waste disposal observation, Waste disposal self-report, Handwashing facility observation, Handwashing facility self-report and Handwashing behavior self-report
Secondary outcomes
Comfort
Sponsors
Name: Dutch Research Council Location: Netherlands Url: https://www.nwo.nl/en/
Partners
Name: James P. Grant School of Public Health, BRAC University Url: https://www.bracu.ac.bd/academics/institutes-and-schools/jpgsph

Cited by

No citation data available for this registration yet.

Keywords