Building Parental Capacity to Help Child Development: A Randomized Controlled Trial of the Save the Children Early Childhood Stimulation Program in Bangladesh 2015, Endline Survey
Bangladesh, 2015
Get Microdata
Reference ID
BGD_2015_SCECSPIE-EL_v01_M
Producer(s)
Marjorie Chinen, Johannes Bos
Metadata
Created on
Mar 06, 2019
Last modified
Mar 06, 2019
Page views
32784
Downloads
2785
Identification
Survey ID Number
BGD_2015_SCECSPIE-EL_v01_M
Title
Building Parental Capacity to Help Child Development: A Randomized Controlled Trial of the Save the Children Early Childhood Stimulation Program in Bangladesh 2015, Endline Survey
Subtitle
Endline Survey
Country/Economy
| Name | Country code |
|---|---|
| Bangladesh | BGD |
Study type
Other Household Survey [hh/oth]
Series Information
Baseline data collection occurred between November 2013 and January 2014, and endline data collection occurred between September and December 2015.
Abstract
Save the Children developed an early stimulation program that delivers actionable messages to mothers and other caregivers that show them how to interact and play with young children.
The program also delivers a Child Development Card and two picture books, and instructions on how to use the card and the books to provide children with early learning opportunities.
The program is low cost and potentially scalable because it builds on an existing delivery platform, and trains current community health care providers to deliver additional messages on early childhood stimulation practices.
The program also delivers a Child Development Card and two picture books, and instructions on how to use the card and the books to provide children with early learning opportunities.
The program is low cost and potentially scalable because it builds on an existing delivery platform, and trains current community health care providers to deliver additional messages on early childhood stimulation practices.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Households
Individuals
Individuals
Scope
Notes
The scope of the Endline Survey for the Impact Evaluation of the Save the Children Early Childhood Stimulation Program includes:
Household Questionnaire : Household profile, deaths in the household; Housing, private cost data questions for mother; Child and health nutrition; Pregnant and lactating mothers; Stimulation knowledge; Decision making/ influence of family members; Responsive feeding; Modified home inventory: infant toddler version; Play materials; Maternal depression; About save the children program (intervention group only); About family welfare assistants household visits; About health assistants with household visits; About visits to community clinics.
Service Provider Questionnaire : Education, experience and training; Working experience; primary tasks; Training and implementation; Distribution of Save the Children materials to beneficiaries; workload; Time spent with households; Perceptions about early childhood developments; Understanding how community clinics operate; Job satisfaction.
Community Leader Questionnaire : Basic infrastructure of the community; Major economic activities; community assets; Migration; About Save the Children Program; External shocks; About the Local Community Clinic.
Household Questionnaire : Household profile, deaths in the household; Housing, private cost data questions for mother; Child and health nutrition; Pregnant and lactating mothers; Stimulation knowledge; Decision making/ influence of family members; Responsive feeding; Modified home inventory: infant toddler version; Play materials; Maternal depression; About save the children program (intervention group only); About family welfare assistants household visits; About health assistants with household visits; About visits to community clinics.
Service Provider Questionnaire : Education, experience and training; Working experience; primary tasks; Training and implementation; Distribution of Save the Children materials to beneficiaries; workload; Time spent with households; Perceptions about early childhood developments; Understanding how community clinics operate; Job satisfaction.
Community Leader Questionnaire : Basic infrastructure of the community; Major economic activities; community assets; Migration; About Save the Children Program; External shocks; About the Local Community Clinic.
Coverage
Geographic Coverage
Bangladesh is divided into seven major administrative regions called divisions, and the study takes place in three of Bangladesh’s seven divisions: Barisal (a southern district), Chittagong (a district in the southeast), and Sylhet (a district in the northeast).
Within these three divisions, the study is located in three districts: Barisal (in the division of Barisal), Chittagong (in the division of Chittagong) and Moulvibazar (in the division of Sylhet).
Districts are subdivided into subdistricts, or upazilas. Within these three districts, the study is located in three upazilas: Muladi (in the district of Barisal), Satkania (in the district of Chittagong), and Kalaura (in the district of Moulvibazar).
Upazilas are subdivided into unions, and the study takes place in 30 unions: 4 unions in Muladi, 16 unions in Satkania, and 10 unions in Kalaura.
Within these three divisions, the study is located in three districts: Barisal (in the division of Barisal), Chittagong (in the division of Chittagong) and Moulvibazar (in the division of Sylhet).
Districts are subdivided into subdistricts, or upazilas. Within these three districts, the study is located in three upazilas: Muladi (in the district of Barisal), Satkania (in the district of Chittagong), and Kalaura (in the district of Moulvibazar).
Upazilas are subdivided into unions, and the study takes place in 30 unions: 4 unions in Muladi, 16 unions in Satkania, and 10 unions in Kalaura.
Universe
Households with children between 3 and 18 months of age residing in the catchment area of participating community clinics at the time of baseline data collection.
Producers and sponsors
Primary investigators
| Name | Affiliation |
|---|---|
| Marjorie Chinen | American Institutes for Research |
| Johannes Bos | American Institutes for Research |
Producers
| Name | Affiliation |
|---|---|
| Rosa Castro-Zarzur | American Institutes for Research |
| Jena Hamadani | International Centre for Diarrhoeal Disease Research in Bangladesh |
| Najmul Hossain | Data International |
| Minhaj Mahmud | BRAC University |
| Matthew Murray | American Institutes for Research |
| Joshua Sennett | American Institutes for Research |
Funding Agency/Sponsor
| Name | Abbreviation | Role |
|---|---|---|
| Strategy Impact Evaluation Fund | SIEF | Funding |
Sampling
Sampling Procedure
Sampling of Households
The study sample frame was generated from community clinic health assistant records, which had the advantage of being the centralized government document of record containing the population frame for all households with children under five years of age. The health assistant dataset included data for all three upazilas of interest. Of a total of 41 unions located in the three upazilas, 11 unions were excluded from the sampling frame. Six of these had incomplete data, and five were excluded because they had only one community clinic and the study design required each union to have at least two clinics. The final sample included 78 community clinics, located in 30 unions.
Within the selected unions and community clinics, eligible households included those with children aged between 3 and 18 months who resided in selected community clinics' catchment areas during the baseline data collection period (November 2013-January 2014).We randomly sampled 33 households from each community clinic's catchment area to participate in the study. The sample was restricted to households with children aged three months or older because the main developmental assessment tool chosen for the evaluation (the Bayley-III; Bayley, 2006) had not been previously validated on children under the age of three months in Bangladesh. Furthermore, because the Bayley-III test is only valid for children up to the age of 42 months, we restricted the upper age limit of participating children to 18 months or younger at the time of baseline data collection in order to collect valid endline data 24 months later.
Replacement
The community clinic health assistant records were not up to date, so the team developed rules for replacing households that were found to be ineligible or "out-of-scope," as well as households that refused to participate. We randomly selected 20 additional replacement households from within each community clinic and included them in a separate list, with each household randomly sorted from 1 to 20. If one of the 33 households originally selected was found to be ineligible or refused to participate, the field interviewer replaced it with the first household from the 20-household replacement list, and continued replacing households in order thereafter.
Overall, the majority of replacements were required because households were identified as ineligible, and only a few replacements were needed for households that refused to participate in the study (N = 39, or 1.5 percent of the sample). Households were ineligible if they did not fit the target sample description: "Households with children from 3-18 months of age that live in the selected community clinics' catchment areas during the period of the baseline data collection." This included: (a) households that had permanently left the catchment area (N = 300); (b) households with incorrect location information in the birth records (N = 291); (c) households with children who were ineligible due to inaccurate birth dates (N =173); and (d) households that were temporarily absent from the catchment area (N =159). For all 39 cases of refusal, the data collectors completed a non-complier questionnaire that captured some basic characteristics of this group to compare with the compliers.
The study sample frame was generated from community clinic health assistant records, which had the advantage of being the centralized government document of record containing the population frame for all households with children under five years of age. The health assistant dataset included data for all three upazilas of interest. Of a total of 41 unions located in the three upazilas, 11 unions were excluded from the sampling frame. Six of these had incomplete data, and five were excluded because they had only one community clinic and the study design required each union to have at least two clinics. The final sample included 78 community clinics, located in 30 unions.
Within the selected unions and community clinics, eligible households included those with children aged between 3 and 18 months who resided in selected community clinics' catchment areas during the baseline data collection period (November 2013-January 2014).We randomly sampled 33 households from each community clinic's catchment area to participate in the study. The sample was restricted to households with children aged three months or older because the main developmental assessment tool chosen for the evaluation (the Bayley-III; Bayley, 2006) had not been previously validated on children under the age of three months in Bangladesh. Furthermore, because the Bayley-III test is only valid for children up to the age of 42 months, we restricted the upper age limit of participating children to 18 months or younger at the time of baseline data collection in order to collect valid endline data 24 months later.
Replacement
The community clinic health assistant records were not up to date, so the team developed rules for replacing households that were found to be ineligible or "out-of-scope," as well as households that refused to participate. We randomly selected 20 additional replacement households from within each community clinic and included them in a separate list, with each household randomly sorted from 1 to 20. If one of the 33 households originally selected was found to be ineligible or refused to participate, the field interviewer replaced it with the first household from the 20-household replacement list, and continued replacing households in order thereafter.
Overall, the majority of replacements were required because households were identified as ineligible, and only a few replacements were needed for households that refused to participate in the study (N = 39, or 1.5 percent of the sample). Households were ineligible if they did not fit the target sample description: "Households with children from 3-18 months of age that live in the selected community clinics' catchment areas during the period of the baseline data collection." This included: (a) households that had permanently left the catchment area (N = 300); (b) households with incorrect location information in the birth records (N = 291); (c) households with children who were ineligible due to inaccurate birth dates (N =173); and (d) households that were temporarily absent from the catchment area (N =159). For all 39 cases of refusal, the data collectors completed a non-complier questionnaire that captured some basic characteristics of this group to compare with the compliers.
Data Collection
Dates of Data Collection
| Start | End | Cycle |
|---|---|---|
| 2015-09-01 | 2015-12-30 | Endline |
Data Collection Mode
Face-to-face [f2f]
Data Collection Notes
Survey instruments were used during baseline data collection, endline data collection, or during the monitoring visits. Baseline data collection occurred between November 2013 and January 2014, and endline data collection occurred between September and December 2015. The monitoring visits occurred in September, October, and December 2014, and during March, May, and July 2015.
Training and Quality Control
Intensive training and piloting took place before data collection at both baseline and endline. The data collection team was divided in two groups: Group 1 focused on the Bayley test, Wolke, and anthropometric measures; while Group 2 focused on household surveys, service providers, and community leaders.
The training for Bayley, anthropometrics, and stimulation practices took approximately six weeks in a centralized location in Dhaka and was led by Dr. Hamadani and her team from ICDDR,B. The training consisted of lectures and discussions, as well as descriptions of the Bayley's manuals and test kits. Participants were divided into groups to perform the tests and observations jointly with the trainers. While a trainee (tester) was administrating the Bayley test, both the trainee and a trainer recorded the observation and the scores. This approach sought to assess and correct scoring gaps between trainers and trainees. Practice sessions continued until enumerators were able to administer the test and observe a child in the presence of a trainer. Five days of training were spent establishing high inter-observer reliability for the Bayley test. Each of the testers took 10 tests to achieve a high degree of reliability between test administrators and trainers. Intra-observer reliability was also assessed through 10 tests of each data collector over another period of five days. For both inter- and intra-observer reliability checks, children from the nearby area were brought to the training venue. In three shifts of testing per day, children were tested in different corners of the training room while trainers observed testers for inter- and intra-observer reliability.
All instruments were piloted prior to both baseline and endline data collection. The team conducted two rounds of pilot testing in order to check the data collection process, protocols, and instruments. The pilot tests helped the team identify and address potential challenges and gave data collectors an opportunity to practice the procedures. The results of the pilot testing led to revised procedures for administering the Bayley-III and anthropometric measures, and revisions to the instruments.
Quality Control During Data Collection
Similar efforts were undertaken to ensure quality during data collection. Several field supervisors ensured that the field enumerators collected reliable and consistent data. They were experienced and familiar with the survey objectives, sampling, and technical and administrative responsibilities. All supervisors remained in the field for the duration of data collection and were responsible for confirming household identities, undertaking spot checks of questionnaires, arranging for suitable testing venues, building rapport with local elected officials, and communicating with the upazila health officer and health assistant before starting the field work.
Training and Quality Control
Intensive training and piloting took place before data collection at both baseline and endline. The data collection team was divided in two groups: Group 1 focused on the Bayley test, Wolke, and anthropometric measures; while Group 2 focused on household surveys, service providers, and community leaders.
The training for Bayley, anthropometrics, and stimulation practices took approximately six weeks in a centralized location in Dhaka and was led by Dr. Hamadani and her team from ICDDR,B. The training consisted of lectures and discussions, as well as descriptions of the Bayley's manuals and test kits. Participants were divided into groups to perform the tests and observations jointly with the trainers. While a trainee (tester) was administrating the Bayley test, both the trainee and a trainer recorded the observation and the scores. This approach sought to assess and correct scoring gaps between trainers and trainees. Practice sessions continued until enumerators were able to administer the test and observe a child in the presence of a trainer. Five days of training were spent establishing high inter-observer reliability for the Bayley test. Each of the testers took 10 tests to achieve a high degree of reliability between test administrators and trainers. Intra-observer reliability was also assessed through 10 tests of each data collector over another period of five days. For both inter- and intra-observer reliability checks, children from the nearby area were brought to the training venue. In three shifts of testing per day, children were tested in different corners of the training room while trainers observed testers for inter- and intra-observer reliability.
All instruments were piloted prior to both baseline and endline data collection. The team conducted two rounds of pilot testing in order to check the data collection process, protocols, and instruments. The pilot tests helped the team identify and address potential challenges and gave data collectors an opportunity to practice the procedures. The results of the pilot testing led to revised procedures for administering the Bayley-III and anthropometric measures, and revisions to the instruments.
Quality Control During Data Collection
Similar efforts were undertaken to ensure quality during data collection. Several field supervisors ensured that the field enumerators collected reliable and consistent data. They were experienced and familiar with the survey objectives, sampling, and technical and administrative responsibilities. All supervisors remained in the field for the duration of data collection and were responsible for confirming household identities, undertaking spot checks of questionnaires, arranging for suitable testing venues, building rapport with local elected officials, and communicating with the upazila health officer and health assistant before starting the field work.
Questionnaires
Questionnaires
Instruments
AIR, ICDDR,B, and Data International Ltd. worked with Save the Children, the World Bank, and the evaluation advisory board to develop the study instruments. The team developed the data collection instruments by drawing from existing national and international tools aligned with the evaluation's outcomes of interest. The core indicators included child development outcomes, anthropometric measures, and parenting stimulation questions, although the final instrument contained many more relevant indicators. Where possible, indicators were measured using questions and approaches that had already been field tested in Bangladesh to ensure that they were appropriate for the local context and the target populations. We also designed the instruments to be of a manageable length in order to avoid interviewer or respondent fatigue and ensure high-quality data. On average, the final survey instruments took 30 minutes to complete.
Endline data collection tools resembled the instruments used at baseline. As discussed above, some instruments were modified slightly based on lessons learned during baseline data collection and monitoring data collection. The non-compliance survey was not administered at endline. Two new measures were added during endline: the Wolke Behavioral Rating Scale, which measures children's behavior during the Bayley-III; and a focus group protocol, with fathers and mothers grouped separately.
AIR, ICDDR,B, and Data International Ltd. worked with Save the Children, the World Bank, and the evaluation advisory board to develop the study instruments. The team developed the data collection instruments by drawing from existing national and international tools aligned with the evaluation's outcomes of interest. The core indicators included child development outcomes, anthropometric measures, and parenting stimulation questions, although the final instrument contained many more relevant indicators. Where possible, indicators were measured using questions and approaches that had already been field tested in Bangladesh to ensure that they were appropriate for the local context and the target populations. We also designed the instruments to be of a manageable length in order to avoid interviewer or respondent fatigue and ensure high-quality data. On average, the final survey instruments took 30 minutes to complete.
Endline data collection tools resembled the instruments used at baseline. As discussed above, some instruments were modified slightly based on lessons learned during baseline data collection and monitoring data collection. The non-compliance survey was not administered at endline. Two new measures were added during endline: the Wolke Behavioral Rating Scale, which measures children's behavior during the Bayley-III; and a focus group protocol, with fathers and mothers grouped separately.
Access policy
Contacts
| Name | Affiliation | |
|---|---|---|
| Shinsaku Nomura | World Bank | snomura@worldbank.org |
Citation requirements
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Example:
Marjorie Chinen, American Institutes for Research, Johanes Bos, American Institutes for Research. Building Parental Capacity to Help Child Development: A Randomized Controlled Trial of the Save the Children Early Childhood Stimulation Program in Bangladesh 2013-2014, Endline Survey. Ref. BGD_2015_SCECSPIE-EL_v01_M. Dataset downloaded from [URL] on [date].
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Example:
Marjorie Chinen, American Institutes for Research, Johanes Bos, American Institutes for Research. Building Parental Capacity to Help Child Development: A Randomized Controlled Trial of the Save the Children Early Childhood Stimulation Program in Bangladesh 2013-2014, Endline Survey. Ref. BGD_2015_SCECSPIE-EL_v01_M. Dataset downloaded from [URL] on [date].
Access authority
| Name | Affiliation | |
|---|---|---|
| Shinsaku Nomura | World Bank | snomura@worldbank.org |
Disclaimer and copyrights
Disclaimer
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.
Metadata production
DDI Document ID
DDI_BGD_2015_SCECSPIE-EL_v01_M_WB
Producers
| Name | Abbreviation | Affiliation | Role |
|---|---|---|---|
| Development Economics Data Group | DECDG | The World Bank | Documentation of the DDI |
Date of Metadata Production
2016-08-09
DDI Document version
Version 01 (August 2016)