Does Survey Section Placement Change the Measurement of Welfare Proxies?

How to cite

Bloem, Jeffrey; Rahman, Wahed. 2022. "Does Survey Section Placement Change the Measurement of Welfare Proxies?." AEA RCT Registry. AEARCTR-0010309. https://doi.org/10.1257/rct.10309-1.2

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

Jeffrey Bloem* , Wahed Rahman

* Primary investigator

Registry ID
AEARCTR-0010309
DOI
https://doi.org/10.1257/rct.10309-1.2
First registered
Last updated
Status
Completed
Trial dates
2022-10-24 – 2023-02-27
Countries
Bangladesh ()
Registry record
http://www.socialscienceregistry.org/trials/10309

Abstract

We will randomly assign survey respondents to two groups. Half of our survey respondents will be assigned to the treatment group (T), for whom the survey section collecting information on household assets will figure early in the survey questionnaire. The other half of our survey respondents will be assigned to the control group (C), for whom the survey section collecting information on household assets will figure late in the survey questionnaire. We will split our respondents evenly between treatment and control groups to maximize statistical power.

Intervention

We will randomly assign survey respondents to two groups. Half of our survey respondents will be assigned to the treatment group (T), for whom the survey section collecting information on household assets will figure early in the survey questionnaire. The other half of our survey respondents will be assigned to the control group (C), for whom the survey section collecting information on household assets will figure late in the survey questionnaire. We will split our respondents evenly between treatment and control groups to maximize statistical power.

Experimental design

We will randomly assign survey respondents to two groups. Half of our survey respondents will be assigned to the treatment group (T), for whom the survey section collecting information on household assets will figure early in the survey questionnaire. The other half of our survey respondents will be assigned to the control group (C), for whom the survey section collecting information on household assets will figure late in the survey questionnaire. We will split our respondents evenly between treatment and control groups to maximize statistical power.

Design details

Randomization method
Randomization done in an office by a computer
Randomization unit
Individual
Clusters
About 6,000 individuals.
Observations
About 6,000 individuals.
Arms
About 3,000 individuals in the treatment group and about 3,000 individuals in the control group.
Primary outcomes
Welfare proxies as measured within the assets module of a household survey

Cited by

No citation data available for this registration yet.

Keywords