Methodological Proposal for the Design and Validation of Research Instruments Supported by Artificial Intelligence

Richar Jacobo Posso Pacheco, Rosangela Caicedo-Quiroz, Giceya Maqueira-Caraballo, Julio Barzola-Monteses, Laura Cristina Barba Miranda and Jorge Ricardo Amancha Gabela

Data and Metadata, 2025, vol. 4, 1103

Abstract: Introduction: the validity of data collection instruments is essential to ensure the quality and replicability of scientific studies; traditional methods require time, resources, and expert participation, making validation difficult. Objective: To develop a procedure for the design and validation of research instruments using Artificial Intelligence as a methodological support tool. Methods: an eight-phase model was designed, ranging from conceptual review and item formulation to linguistic evaluation, simulated rational validation, comprehension verification, internal consistency analysis, and structural optimization. Results: the process demonstrated applicability, technical coherence, and practical utility. ChatGPT 4.5 enabled the automation of analyses and the generation of content aligned with theoretical constructs, optimizing the preliminary validation phases. Conclusions: AI represents a viable alternative in resource-limited settings. While it does not replace classic empirical methods, it complements methodological rigor in key stages. Ethical and technical protocols must be established for its responsible use in scientific research.

Date: 2025
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:dbk:datame:v:4:y:2025:i::p:1103:id:1056294dm20251103

DOI: 10.56294/dm20251103

Access Statistics for this article

More articles in Data and Metadata from AG Editor
Bibliographic data for series maintained by Javier Gonzalez-Argote ().