Description
|
We present the analysis dataset, script, as well as the data collection forms from an evaluation of an 11-week, remote early learning program (RELP) delivered via WhatsApp calls and messages to families in areas of Lebanon with little or no access to ECE. The program was created in response to COVID-19 with a focus on Syrian refugee and vulnerable Lebanese families and tested in hard-to-access areas in order to capture impacts on those most likely to need remote ECE post-pandemic. It is shorter than typical school programming because in under-resourced, conflict-affected, and displaced settings it is not always possible to find the political will and resources, or to maintain families, for longer programming (see Bonilla et al., 2019). Thus, in seeking to provide ECE in such settings, it is critical that we understand the impact of programs that are shorter, cheaper, and potentially more scalable than full-year, in-person options. In this evaluation, we examine the impacts of both i) RELP alone and ii) RELP provided alongside a remote parenting support program (RPSP). Impacts presented are in comparison to a wait list control group who received RELP immediately following endline data collection. Analyses examine experimental impacts on child development, parenting, and caregiver well-being. The files provided include: 1) The analysis dataset titled "RELP_clean_deidentified". This dataset includes all outcome variables as well as covariates. Use "Models_dofile" to replicate the analyses. 2) The STATA .do file used to run the analyses, titled "Models_dofile" 3) The CAPI xls forms that were used to collect baseline and endline data for both the IDELA measure of child development as well as the caregiver surveys (baseline and endline) 4) The STATA .do file used to clean the data, run psychometrics and obtain outcome scores, and run the multiple imputation. This file is called "RELP_raw_to_analytic" and it is meant to provide transparency around data transformations that took place starting with the raw data and ending with the analysis dataset "RELP_clean_deidentified". 5) A pdf copy of the published pre-registration document. The study and analysis plan were pre-registered in REES (ID:13920). (2024-03-30)
|