Estimating Enumerator Effects: An Experimental Analysis in West Africa

How to cite

Doces, John. 2021. "Estimating Enumerator Effects: An Experimental Analysis in West Africa." AEA RCT Registry. AEARCTR-0008194. https://doi.org/10.1257/rct.8194-1.0

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

John Doces*

* Primary investigator

Registry ID
AEARCTR-0008194
DOI
https://doi.org/10.1257/rct.8194-1.0
First registered
Last updated
Status
On going
Trial dates
2021-06-07 – 2021-09-20
Countries
Côte d'Ivoire (Abidjan)
Registry record
http://www.socialscienceregistry.org/trials/8194

Abstract

This analysis will estimate the effects of enumerators on wellbeing in Africa. Based on an existing exploratory analysis, this second part of the paper serves as a replication to the initial analysis which looked to determine if there were enumerator effects. The initial results indicate that there are enumerator effects. The purpose in this second round of replication is to see if there is a difference in responses to the question on wellbeing based on the nationality of enumerator (African versus American), and to test for manipulation checks. Thus, this second round will specify a hypotheses based on the first round of data and then test that hypothesis with new data from a new cluster.

Intervention

Primary intervention is the enumerator. Randomly selected respondents will conduct survey with American enumerator (treatment group) and other respondents will be randomly assigned a local, African enumerator (control group).

Experimental design

The experimental design is a simple randomized experiment in which respondents will be treated with an American enumerator to see how their responses vary to a question about their personal wellbeing. The control is a local African enumerator. This part of the project is a replication of an on-going study that was initiated to see if there was a treatment effect. Thus, it represents a replication of that study but now with a specific hypothesis and it will also test for treatment effects. Both enumerators will be the same from the original study. Sample will be an entirely different neighborhood in Abidjan than in the prior study.

The hypothesis being tested is as follows:

In a comparison of individuals, those treated by the American enumerator will report higher levels of wellbeing than those treated with the African enumerator.

Design details

Randomization method
Randomization will be done in the field. Each enumerator will begin on different streets in a densely populated lower-income neighborhood of Abidjan. Respondents will be randomly selected according to the flip of a coin using a predetermined randomly generated lists of 0s and 1s.
Randomization unit
Randomization will be at the individual level.
Clusters
There will be one neighborhood in Abidjan.
Observations
Power analysis using previously collected data with the same treatment and outcome variables in the Ivory Coast indicate at a power of 0.80 that 124 total observations or 62 per group will need to be collected. For purposes of this study, we will collect a total of 70 per respondent for a total of 140 individuals. With 140 observations, the power is 0.85.
Arms
There will be one cluster with 70 individual respondents per treatment.
Primary outcomes
Primary outcome is a measure of wellbeing based on the Cantril index. The measure is as follows: Please imagine a ladder with steps numbered from zero (0) at the bottom to ten (10) at the top. The top of the ladder represents the best possible life for you and the bottom of the ladder represents the worst possible life for you. On which step of the ladder would you say you personally feel you stand at this time?
Secondary outcomes
There will be three additional questions to test for the treatment effects. -How close do you feel to foreigners like Americans? (0 = not close at all, 1 = somewhat close, 2= close, 3= very close). -Do you think wellbeing in the Ivory Coast is better, worse, or about the same as in the United States? (Better, Worse, Same) -Compared to the United States, how powerful economically and militarily is the Ivory Coast in your opinion? (0 = not powerful at all, 1 = somewhat powerful, 2 = powerful, 3 = very powerful

Cited by

No citation data available for this registration yet.

Keywords