DIVE-PL

A reproducible AI tool-chain methodology for building survey research studies.

Related work

Each component has prior art. The novelty is the integration of all four plus a full open-science reproducibility layer, demonstrated on a real prospective human survey study. Literature scan: mid-2026 (preprints flagged).

ComponentNearest precedentEstablishedOur difference
LLM authors instrumentCurrent Psychology 2025; EJPA 2024; PLOS ONE 2026; J. Patient-Rep. Outcomes 2025LLM-generated scales reproduce factor structure & reliability of gold-standard instrumentsAuthoring embedded in an EDC + reviewer + dashboard + open-science chain
Second LLM as reviewerarXiv:2501.05985 (2025); ChatGPTest arXiv:2405.06329Generator+reviewer loop and LLM critique of item wording existReviewer used as a formal, logged QC gate feeding registration-grade instruments
REDCap + LLMKaiyrbekov et al., JAMIA Open 2025; Vanderbilt/Stanford modules (grey lit)LLM+REDCap capture works; AI-built REDCap instruments mostly grey litAI-assisted construction of registration-ready REDCap instruments, documented
AI-built dashboardsInformatics (MDPI) 2026; REDCap+dashboard+chatbot PMC12099335 (2025)Generative-AI ETL/visualisation as methodology existsDashboard published openly as part of the reproducibility layer
Conceptual AI survey frameworkFrontiers in Digital Health 2025A conceptual framework existsWe provide an implemented, end-to-end, reproducible pipeline + case study

One-line verdict. Every brick is published; the building is not. DIVE-PL is a methods/framework paper with an implementation case study, citing precedents honestly.

Key references

  1. Exploring LLMs for Automated Generation and Adaptation of Questionnaires — arXiv:2501.05985
  2. ChatGPTest: questionnaire pretesting with GPT — arXiv:2405.06329
  3. Automated survey collection with LLM agents — JAMIA Open 2025
  4. Generative-AI-assisted clinical data processing & dashboards — Informatics (MDPI) 2026
  5. Rethinking survey development in health research with AI — Frontiers in Digital Health 2025
  6. REDCap + dashboards + AI chatbot architecture — PMC12099335 (2025)
  7. Automated psychology questionnaire generation with LLMs — PLOS ONE 2026