Avaliação da experiência do usuário na utilização de um protótipo de chatbot para o apoio à prevenção do câncer do colo do útero

Luiz Fernando Picolo; Manuella Silva Sampaio; Felipe de Mauro Paulo; Renato de Souza Garcia

doi:10.1590/SciELOPreprints.18459

##article.authors##

Luiz Fernando Picolo Instituto Federal de Educação Ciência e Tecnologia do Mato Grosso do Sul https://orcid.org/0000-0002-0768-8691
- Writing – Original Draft Preparation
- Conceptualization
- Data Curation
- Formal Analysis
- Investigation
- Software
- Supervision
- Visualization
Manuella Silva Sampaio Instituto Federal de Educação Ciência e Tecnologia do Mato Grosso do Sul https://orcid.org/0009-0005-7814-7648
- Conceptualization
- Methodology
- Supervision
- Writing – Review & Editing
Felipe de Mauro Paulo Instituto Federal de Educação Ciência e Tecnologia do Mato Grosso do Sul https://orcid.org/0009-0004-2879-3329
- Conceptualization
- Methodology
- Validation
- Writing – Review & Editing
Renato de Souza Garcia Instituto Federal de Educação Ciência e Tecnologia do Mato Grosso do Sul https://orcid.org/0009-0008-9811-6058
- Data Curation

DOI:

https://doi.org/10.1590/SciELOPreprints.18459

Keywords:

Conversational agents , Cervical cancer, Digital health , User experience , Usability

Abstract

This study evaluated the usability, acceptance, and flow experience associated with interaction with the Liz chatbot prototype, developed to support cervical cancer prevention in the public healthcare system of Nova Andradina, Mato Grosso do Sul, Brazil. A mixed-methods study was conducted with 58 participants aged 18–64 years, using the System Usability Scale (SUS), Technology Acceptance Model (TAM), and Flow State Scale short version (FSS-short), as well as an open-ended question. Data were analyzed using instrument scores and lexical analysis with IRAMUTEQ software. The SUS score was 75.5, indicating good usability. The TAM results revealed favorable perceptions of usefulness, ease of use, and intention to use, while the FSS-short indicated good participant engagement during the interaction. In the textual analysis, the term scheduling stood out, associated with ease of use, clarity, and functionality. However, occasional difficulties with typing, chatbot responses, and errors were also reported. The findings indicate a positive initial evaluation and highlight the contribution of integrating standardized assessment measures with users’ perceptions to improve conversational interfaces in healthcare.