Rapid, Tailored Dietary and Health Education Through A Social Media Chatbot Microintervention: Development and Usability Study With Practical Recommendations
Bibliographic record
Abstract
BACKGROUND: There is an urgent need to innovate methods of health education, which can often be resource- and time-intensive. Microinterventions have shown promise as a platform for rapid, tailored resource dissemination yet have been underexplored as a method of standardized health or dietary education; social media chatbots display unique potential as a modality for accessible, efficient, and affordable educational microinterventions. OBJECTIVE: This study aims to provide public health professionals with practical recommendations on the use of social media chatbots for health education by (1) documenting the development of a novel social media chatbot intervention aimed at improving dietary attitudes and self-efficacy among South Asian American young adults and (2) describing the applied experiences of implementing the chatbot, along with user experience and engagement data. METHODS: In 2023, the "Roti" chatbot was developed on Facebook and Instagram to administer a 4-lesson tailored dietary health curriculum, informed by formative research and the Theory of Planned Behavior, to 18- to 29-year-old South Asian American participants (recruited through social media from across the United States). Each lesson (10-15 minutes) consisted of 40-50 prescripted interactive texts with the chatbot (including multiple-choice and open-response questions). A preintervention survey determined which lesson(s) were suggested to participants based on their unique needs, followed by a postintervention survey informed by the Theory of Planned Behavior to assess changes in attitudes, self-efficacy, and user experiences (User Experience Questionnaire). This study uses a cross-sectional design to examine postintervention user experiences, engagement, challenges encountered, and solutions developed during the chatbot implementation. RESULTS: Data from 168 participants of the intervention (n=92, 54.8% Facebook; n=76, 45.2% Instagram) were analyzed (mean age 24.5, SD 3.1 years; n=129, 76.8% female). Participants completed an average of 2.6 lessons (13.9 minutes per lesson) and answered an average of 75% of questions asked by the chatbot. Most reported a positive chatbot experience (User Experience Questionnaire: 1.34; 81/116, 69.8% positive), with pragmatic quality (ease of use) being higher than hedonic quality (how interesting it felt; 88/116, 75.9% vs 64/116, 55.2% positive evaluation); younger participants reported greater hedonic quality (P=.04). On a scale out of 10 (highest agreement), participants reported that the chatbot was relevant (8.53), that they learned something new (8.24), and that the chatbot was helpful (8.28). Qualitative data revealed an appreciation for the cheerful, interactive messaging of the chatbot and outlined areas of improvement for the length, timing, and scope of text content. Quick replies, checkpoints, online forums, and self-administered troubleshooting were some solutions developed to meet the challenges experienced. CONCLUSIONS: The implementation of a standardized, tailored health education curriculum through an interactive social media chatbot displayed strong feasibility. Lessons learned from challenges encountered and user input provide a tangible roadmap for future exploration of such chatbots for accessible, engaging health interventions.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.015 | 0.014 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".