Disentangling Channels of Information Dissemination in Healthcare Delivery in Rural India

How to cite

Berg, Erlend; D, Rajasekhar; R, Manjula; Roy, Sanchari. 2023. "Disentangling Channels of Information Dissemination in Healthcare Delivery in Rural India." AEA RCT Registry. AEARCTR-0012243. https://doi.org/10.1257/rct.12243-1.0

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

Erlend Berg* , Rajasekhar D , Manjula R , Sanchari Roy

* Primary investigator

Registry ID
AEARCTR-0012243
DOI
https://doi.org/10.1257/rct.12243-1.0
First registered
Last updated
Status
In development
Trial dates
2023-09-01 – 2024-03-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/12243

Abstract

We describe the analysis plan for a field experiment that will examine the importance of caste and salience in the transmission of information in rural India. The participants are social workers and potential beneficiaries living in randomly selected villages in Karnataka. The social workers are provided information about a particular health insurance scheme implemented by the Indian government. We then ask them to transmit this information to the potential beneficiaries. Each social workers is matched with four potential beneficiaries, and we measure information transmission by fielding a knowledge test to the beneficiaries before and after their interaction with the social worker. We randomly vary whether or not social worker and beneficiaries are matched on caste, and whether or not beneficiary caste is revealed to the social worker.

Intervention

Every ASHA is matched with two beneficiaries from her own caste group and two from a different caste group. Each beneficiary will interact with only one ASHA. We experimentally vary whether the beneficiary’s caste identity is revealed to the ASHA or not.

Hence, there will be four experimental groups: ASHA--beneficiary pairs who belong to same caste and beneficiary’s caste identity is revealed (Matched-Revealed), pairs who are matched but identity is not revealed (Matched-Not revealed), pairs who are unmatched and identity is revealed (Unmatched-Revealed) and pairs who are unmatched but identity is not revealed (Unmatched-not revealed). Each ASHA will meet with one household from each of these four categories, and the order in which she interacts with them will be randomised.

Experimental design

Every ASHA is matched with two beneficiaries from her own caste group and two from a different caste group. Each beneficiary will interact with only one ASHA. We experimentally vary whether the beneficiary’s caste identity is revealed to the ASHA or not.

Hence, there will be four experimental groups: ASHA--beneficiary pairs who belong to same caste and beneficiary’s caste identity is revealed (Matched-Revealed), pairs who are matched but identity is not revealed (Matched-Not revealed), pairs who are unmatched and identity is revealed (Unmatched-Revealed) and pairs who are unmatched but identity is not revealed (Unmatched-not revealed). Each ASHA will meet with one household from each of these four categories, and the order in which she interacts with them will be randomised.

Design details

Randomization method
Computer.
Randomization unit
ASHA-beneficiary pair. Each ASHA is allocated two matched and two unmatched households. Caste revelation is randomised with these pairs.
Clusters
400 ASHA-beneficiary pairs
Observations
400 ASHA-beneficiary pairs
Arms
100 matched-revealed ASHA-beneficiary pairs 100 matched-notrevealed ASHA-beneficiary pairs 100 unmatched-revealed ASHA-beneficiary pairs 100 unmatched-notrevealed ASHA-beneficiary pairs
Primary outcomes
beneficiary knowledge of the health insurance scheme
Secondary outcomes
Possible secondary outcome variables: Measures of 'teaching to the test', subjective perception of counterparty friendliness and communication effectiveness. Possible secondary outcome variables: Measures of 'teaching to the test', subjective perception of counterparty friendliness and communication effectiveness. It will also be of interest to look at the effect for sub-categories of knowledge, for example, numerical vs non-numerical knowledge, and eligibility vs information on benefits.

Cited by

No citation data available for this registration yet.

Keywords