ParentText Optimisation Study: A Chatbot-led Intervention Aimed to Promote Playful Parenting and Prevent Violence Against Children

Project outline

Background

In response to the increased risk of violence against children due to the COVID-19 pandemic and existing barriers to taking parenting programmes to scale in LMICs, the Parenting for Lifelong Health (PLH) initiative adapted its parenting programmes into a self-guided chatbot called ParentText. While there is increasing evidence of the effectiveness of digital parenting programmes as a potentially scalable alternative to in-person programme delivery, enrolment and engagement rates remain low. Furthermore, this evidence is limited to high-income countries, with no studies conducted on parenting chatbot interventions nor any digital programmes in low- and middle-income countries (LMICs). 
Intervention 

ParentText delivers playful parenting and violence prevention content to parents/caregivers of children ages 0 to 17 years via interactive text messages over five to 12 weeks. Programme content includes modules on positive parenting, learning through play, behaviour management, online safety, sexual abuse prevention, partner relationships, and stress reduction. The chatbot also includes components aimed at increasing user engagement, such as gender-targeting, personalisation, and gamification.

Objectives 

This research has five research objectives. First, we want to examine the acceptability and feasibility of ParentText. Second, we aim to qualitatively identify the barriers and facilitators to enrolment and engagement. Third, we aim to optimise ParentText so that we can implement the most effective enrolment and engagement strategy, to achieve the highest possible retention and engagement rate. Fourth, we aim to examine the preliminary effect of ParentText on improving playful parenting and reducing the risk of violence again children. Fifth, we want to examine whether there are different levels of enrolment, engagement and effectiveness based on different population characteristics. Due to the multi-dimensional nature of enrolment and engagement, a mixed-method approach will be implemented to address these research aims.

Study Design 

The study will be conducted in Jamaica, Malaysia, Philippines, South Africa and Sri Lanka. Adult participants will be primarily recruited through UNICEF, government, and NGO platforms. The study consists of three phases based on the Preparation and Optimsiation phases of the Multiphase Optimisation Study framework (Collins, 2018). Phase 1 uses a mixed-method approach to examine the feasibility, engagement, and acceptability of the intervention. Phase 2 uses four successive factorial experiments to optimise recruitment, engagement, and retention based on the type, targeting, and frequency of messages. Phase 3 will examine the effect of the intervention over time with assessments conducted every other week measuring positive parenting, emotional and physical abuse, child behaviour, parental self-efficacy, parental stress, sexual abuse prevention, attitudes towards gender roles, and intimate partner violence. Analyses will include thematic analysis (Phase 1) regression models (Phase 2), and latent growth models examining change over time (Phase 3).