Can technology narrow the early childhood stimulation gap in rural Guatemala? Results from an experimental approach

The intervention consists of an early stimulation program through the radio. Voice messages will be used to incentivize the take up of the intervention and reinforce parental-child interactions. Specifically, we will have three treatment arms:
(T1 ) radio theater skit alone + digital cash (or in-kind) transfers ($40 value only once),
(T2) T1 + recorded voice message to build awareness about the radio intervention and incentivize take-up
(T3) T2 + phone message about how mother/caregiver could interact with her child.

2020-06-30

2020-09-30

1. Take up of radio messages
2. Self-reported parental behavior on parent -child interactions (Play activities sub scale of the Family Care Indicators
3. Maternal anxiety (General Anxiety Disorder (GAD-7))
4.Child development using two instruments: (1) Caregiver Reported Early Childhood Development Instrument-Short Form (CREDI-SF) to measure child overall development status, and (2) MacArthur-Bates Communicative Development Inventories Short Form (MA-CDI-SF) to measure language skills

FAMILY CARE INDICATORS (FCI; from Hamadani et al 2010)

Play Activities Subscale
In the past week, on how many days did you or any other adult family member do the following
with (CHILD)? (read list; indicate ‘yes’ or ‘no’ for each)

Read books or look at picture-books with child
Tell stories to child
Sing songs with child
Take child outside home place
Play with the child with toys
Spend time with child in naming things, counting, drawing


ANXIETY: GAD-7
Instruction: Over the last 2 weeks, how often have you been bothered by the following problems?
Variable Items
gad1 1. Feeling nervous, anxious or on edge.
gad2 2. Not being able to stop or control worrying.
gad3 3. Worrying too much about different things.
gad4 4. Trouble relaxing.
gad5 5. Being so restless that it is hard to sit still.
gad6 6. Becoming easily annoyed or irritable.
gad7 7. Feeling afraid as if something awful might happen.

response options: 0 = Not at all
1 = Several days
2 = More than half the days
3 = Nearly every day

Caregiver Reported Early Childhood Development Index (CREDI) - Consists of 20 questions by age group (0-5 months, 6-11 months, 12-17 months, 18-23 months, 24 to 29 months, 30-35 months). Examples of questions for the first age group are:

CREDI_A1 Does the child smile when others smile at him/her?
CREDI_A2 Does the child grasp onto a small object (e.g., your finger, a spoon) when put in his/her hand?
CREDI_A3 Does the child recognize you or other family members (e.g., smile when they enter a room or move toward them)?
CREDI_A4 Does the child show interest in new objects by trying to put them in his/her mouth?
CREDI_A5 When lying on his/her stomach, can the child hold his/her head and chest off the ground using only his/her hands and arms for support?
CREDI_A6 Can the child pick up a small object (e.g., a small toy or small stone) using just one hand?
CREDI_A7 When lying on his/her back, does the child grab his/her feet?
CREDI_A8 Does the child look at an object when someone says "look!" and points to it?
CREDI_A9 Does the child look for an object of interest when it is removed from sight or hidden from him/her (e.g., put under a cover, behind another object)?
CREDI_A10 Does the child intentionally move or change his/her position to get objects that are out of reach?
CREDI_A11 Does the child play by tapping an object on the ground or a table?
CREDI_A12 Can the child hold him/herself in a sitting position without help or support for longer than a few seconds?
CREDI_A13 Can the child pick up and eat small pieces of food with his/her fingers?
CREDI_A14 Can the child transfer a small object (e.g., a small toy or small stone) from one hand to the other?
CREDI_A15 Can the child use gestures to indicate what he/she wants (e.g., put arms up to indicate that he/she wants to be held, or point to water)?
CREDI_A16 Can the child crawl, roll, or scoot forward on his/her own?
CREDI_A17 Can the child throw a small ball or small stone in a forward direction using his/her hand?
CREDI_A18 Can the child pick up and drop a small object (e.g., a small toy or small stone) into a bucket or bowl while sitting?
CREDI_A19 Can the child say one or more words (e.g., names like "Mama" or "ba" for "ball")?
CREDI_A20 Can the child walk several steps while holding on to a person or object (e.g., wall or furniture)?

Response options are: Yes, No, I don't know.

The intervention consists of an early stimulation program through the radio. Voice messages will be used to incentivize the take up of the intervention and reinforce parental-child interactions. Specifically, we will have three treatment arms:

(T1)radio theater skit alone + digital cash transfers (CT) ($40 only once),
(T2) T1 + recorded voice message to build awareness about the radio intervention and incentivize take-up
(T3) T2 + phone message about how mother/caregiver could interact with her child.

Randomization of intervention arms (T1, T2, T3) will ensure balance between groups in terms of observable and unobservables characteristics. The evaluation will measure the value added of including awareness campaigns to the radio intervention and the value of combining awareness campaign with child development messages on top of radio. [T1 is our experimental control]

The intervention consists of an early stimulation program through the radio. Voice messages will be used to incentivize the take up of the intervention and reinforce parental-child interactions. Specifically, we will have three treatment arms:

(T1)radio theater skit alone + digital cash transfers (CT) ($40 only once),
(T2) T1 + recorded voice message to build awareness about the radio intervention and incentivize take-up
(T3) T2 + phone message about how mother/caregiver could interact with her child.

Randomization of intervention arms (T1, T2, T3) will ensure balance between groups in terms of observable and unobservables characteristics. The evaluation will measure the value added of including awareness campaigns to the radio intervention and the value of combining awareness campaign with child development messages on top of radio. [T1 is our experimental control]

Analyzing take up of comparison group will inform on the take up rate in the absence of other interventions. Our comparison group is composed of children under age 3, living in communities that won’t have access to any of the treatments, living in similar communities in rural areas, with similar characteristics. These children live in areas where ChildFund operates. We will use propensity score matching techniques to guarantee that these children are similar in observed characteristics to the treatment group.

Randomization done in office by a computer

Unit of randomization is community, but we are collecting data at the individual level. Thus, individuals are clustered in a community.

Yes

35 clusters (communities)

29 observations per cluster 35 clusters x 29 families per cluster x 3 treatment arms=1,015 x 3= 3,045 families

35 communities in treatment 1, 35 communities in treatent 2, 35 clusters in treatment 3

The calculation of MDE were done with response rates instead of total sample. I expect 20% attrition and 90% take up. Thus, I expect to collect data for 20 families per cluster MDE outcome Family care indicator - play activity scale=0.32 (mean=2.53, SD=1.27, power=.80) MDE outcome General anxiety disorder =0.12 (mean=30%, power=.80). MDE outcome MacArthur-Bates Communicative Development Inventory = 7.54 (mean=54.01, SD=29.88, power=.80) MDE outcome Caregiver-Reported Early Development Instrument= 0.06 (mean=0.2 [normalized], SD=0.25, power=0.80)