Evaluating Multilingual Early Reading as the Groundwork for Education: A Cluster-Randomized Trial in Kenya

How to cite

Ozier, Owen; Fernald, Lia; Knauer, Heather; Jakiela, Pamela. 2019. "Evaluating Multilingual Early Reading as the Groundwork for Education: A Cluster-Randomized Trial in Kenya." AEA RCT Registry. AEARCTR-0004425. https://doi.org/10.1257/rct.4425-1.2000000000000002

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

Owen Ozier* , Lia Fernald , Heather Knauer , Pamela Jakiela

* Primary investigator

Registry ID
AEARCTR-0004425
DOI
https://doi.org/10.1257/rct.4425-1.2000000000000002
First registered
Last updated
Status
Completed
Trial dates
2017-01-01 – 2020-01-01
Countries
Kenya ()
JEL codes
I25, O15
Registry record
http://www.socialscienceregistry.org/trials/4425

Abstract

Encouraging Multilingual Early Reading as the Groundwork for Education (EMERGE) is a cluster-randomized evaluation of a program that provides Kenyan parents with illustrated children’s storybooks and modified dialogic reading training. The ongoing study employs a 73-community cluster-randomized design to examine the impact of a cost-effective parent education and storybook distribution program on preschool-aged children in rural Kenya. Within this main design, we use a within-cluster caregiver-level randomization to examine the impact of the language of the children's storybooks.

Intervention

Caregivers (typically mothers, fathers, or grandmothers) receive a two-to-three-hour modified dialogic reading training (also referred to as “book-sharing” – conveying the idea that reading to/with a young child can be an interactive process); they also receive a bundle of five illustrated storybooks. More details can be found in the baseline report, which is provided as an external link in the main trial information.

Experimental design

The main EMERGE study is a cluster-randomized trial. Our research design, a two-stage randomization implemented in 73 rural communities in Nyando sub-county, Kenya. First, communities were randomly assigned to either the treatment group or the control group. Within treatment communities, caregivers were further randomized into two treatment arms: Luo-language storybooks or English-language storybooks. All caregivers (who completed the baseline) in treatment villages were invited to attend the dialogic reading training, which was conducted in Luo. Books were distributed after the training according to the caregiver-level randomization.

Design details

Randomization method
Stratified random assignment using Stata 14.2.
Randomization unit
Two-stage randomization: community-level and household-level within treatment communities
Clusters
73 communities, 2,013 households/caregivers
Observations
2,013 caregivers with ~2,500 children who were assessed at baseline and ~2,500 additional children (within the same households).
Arms
COMMUNITY-LEVEL RANDOMIZATION (STORYBOOKS VS. CONTROL): Control group: 37 communities, 997 caregivers, 1,260 children assessed at baseline. Treatment group: 36 communities, 1,016 caregivers, 1,267 children assessed at baseline. HOUSEHOLD-LEVEL RANDOMIZATION WITHIN TREATMENT COMMUNITIES (LUO STORYBOOKS VS. ENGLISH STORYBOOKS): Luo books: 508 caregivers, 635 children assessed at baseline. English books: 508 caregivers, 632 children assessed at baseline.
Primary outcomes
child vocabulary, child literacy skills, reading/book-sharing frequency
Secondary outcomes
child stimulation, caregiver time use, parental beliefs about children's abilities, parental self-efficacy

Cited by

No citation data available for this registration yet.

Keywords