The “Digital Plus” Approach to Upscaling Early Childhood Development Services: Evaluating the Effectiveness of a Mixed Online-Offline Caregiver Education Program in Rural China

Past research has established that home visit-based parent education programs are effective in promoting cognitively stimulating parenting practices and early child development in households with children under the age of 2 years. The objective of the current study is to determine whether mixed offline and online delivery of parenting trainings can maintain intervention effectiveness while reducing costs and increasing scalability. While parenting trainings will be conducted via in-person home visit during the first two months of the intervention, most trainings beyond this point will occur via prerecorded video through a specially designed app, reducing the number of trainers needed to deliver the trainings throughout the study period. On the village level, the intervention will be implemented by two types of municipal government-affiliated authorities overseeing women’s and family issues in rural areas of China, peer mothers (tongban mama) and children’s committee directors (ertong zhuren). Moreover, local doctors will be asked to send reminder texts to caregivers in the treatment group to remind them to engage with online training materials.

By reducing the number of staff needed overall to conduct trainings and entrusting most village-level implementation to local authorities in the municipal and medical systems, this project design has the potential to be significantly upscaled or expanded to remote reasons with relatively low additional cost. We also take various steps to offset potential negative impacts of offline delivery on the trainings’ effectiveness in promoting ECD, including conducting additional in-home visits every two months during the primarily online delivery period. Therefore, we hypothesize that this mixed online-offline study design will produce significantly positive effects on ECD outcomes comparable to those found in past, fully-offline parenting training interventions.

2024-08-15

2025-08-15

The primary outcome of this study is early childhood psychosocial development. This outcome will be measured at baseline and endline (i.e., after 1 year at the end of the trial) using the following scales:
1. Bayley Scales of Infant and Toddler Development, third edition (Bayley-III)
2. Ages and Stages Questionnaires: Social-emotional (ASQ:SE)
3. New Wolke Social-Emotional Behavior Ratings

Secondary outcomes to be measured at baseline and endline (i.e., after 1 year at the end of the trial) include the following:

1. Parenting and attachment style, measured with:
1.1. The World Value Survey
1.2 The Mother’s Object Relations Scale – Short Form (MORS-SF)

2. Parenting stress, measured with the Parenting Stress Index

3. Parental investments in stimulating parenting practices and materials, measured with:
3.1. The Family Care Indicators (FCI)
3.2. Home Observation Measurement of the Environment – Short Form (HOME-SF)

4. Parenting self-efficacy, measured with the TOol to measure Parenting Self-Efficacy (TOPSE)

5. Caregiver mental health, measured using:
5.1. The Patient Health Questionnaire (PHQ)
5.2. The Depression, Anxiety, and Stress Scales, 21 items (DASS-21)

6. Positive perception of daily chores, measured with the Parenting Daily Hassles scale (PDH)

7. Structural and functional social support (community/family), measured with the Multidimensional Scale of Perceived Social Support (MSPSS)

8. Caregiver digital literacy, measured using a self-designed questionnaire

9. Caregiver and child screen exposure

The Digital Plus project seeks to develop, deliver, and evaluate a multi-stage, mixed-format caregiver education program to target early childhood psychosocial development in rural China. The intervention contains two arms, one that receives the early childhood development education services (treatment) and one that does not (control), and will occur in two phases.

The first phase is fully offline and will last two months. During this phase, caregivers in the treatment group will receive weekly in-home, in-person lessons from parenting trainers recruited for this project. The content of the lessons mostly consists of an activity-based parenting training curriculum that is based the preexisting Reach Up and Learn curriculum and was adapted by the research team with assistance from early childhood development experts in China. In the adapted curriculum, parenting trainers will conduct one session per week that guide caregivers through interactive activities that are designed to stimulate the early development of children between 6 and 36 months old across four dimensions: cognitive, language, motor, and social-emotional skill development. Trainers will also lend caregivers toys and books and instruct them to use these materials to practice that week’s activities before the next session. In addition to the activity-based curriculum, trainers will also deliver feedback about the caregiver’s interactions with the target child using a handbook developed by the research team in partnership with experts from China.

The second phase contains mixed online and offline deployment and will last ten months. During this phase, caregivers will continue to receive activity-based lessons following the parenting curriculum, but the lessons will take the form of prerecorded videos delivered through an app developed for this project. Community health workers will continue to lend caregivers toys and books related to each week’s lesson and will be on hand to assist caregivers if they have questions. Once every two months, the parenting trainers will conduct another home visit to answer questions and offer feedback on the caregiver’s interaction skills using the handbook.

Not available

Randomization done in office by a computer

Village

Yes

80 villages

1280 caregiver-child dyads

40 villages control, 40 villages treatment

Power was estimated based on the treatment effect of parenting interventions documented in the literature (Cunha, Gerdes, and Nihtianova n.d.; Emmers et al., 2021) between study arms for two primary outcome indicators: overall child development and active parenting practices. Our sample size estimates accounted for randomization at the cluster level with a 0.05 significance level. The impact size of the parenting intervention is scaled back to 0.2 standard deviation (SD) on child development and 0.3 SD on active parenting practices by arbitrary assumption. Overall child development. The power of the design is calculated given a 0.2 SD effect assuming an intra-cluster correlation of 0.01 and a cluster size of 16 babies per village. Active Parenting Practices. The power of the design is calculated given a 0.3 SD effect assuming an intra-cluster correlation of 0.01 and a cluster size of 16 babies per village. Based on these reference effect sizes, our study design holds powers of 0.91 and 0.99 when overall child development and active parenting practices serve as the outcome variable, respectively. To detect the difference in treatment effect between treatment arms, the minimal detectable effect needs to be 0.17 standard deviations (power = 0.8).