RS/AT: uma abordagem para aplicação de análise temática em revisão sistemática

• Ano de defesa: 2016
• Autor(a) principal: Gastaldi, Rafael Adabo
• Orientador(a): Fabbri, Sandra Camargo Pinto Ferraz
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Português: Estudo secundário; Revisão sistemática; Análise qualitativa de dados; Análise temática; Engenharia de software; Start; Insight
• Palavras-chave em Inglês: Secondary study; Systematic review; Qualitative analysis; Thematic analysis; Software engineering
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::METODOLOGIA E TECNICAS DA COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/10540

Context: Systematic reviews have been increasingly applied in computing area and are used to perform a broad literature search and generate the state of the art, or evidences on an issue. For this, primary studies are identified, selected and entirely read, leading to a hard work of reading and analysis. Thematic analysis is a way to analyze data qualitatively, through the identification of relevant information that are grouped into categories. Aim: Present the RS/AT approach that applies thematic analysis in the context of systematic reviews, with computational support, aiming to extract relevant data from the literature gathered in the review. Method: Once defined how to use thematic analysis in the systematic review process, the computational support was developed based on the StArt tool (which supports systematic reviews) and the Insight tool (which supports the qualitative analysis and, therefore, the thematic analysis). Experimental studies were conducted to evaluate the feasibility of the approach application and possible contributions to the data extraction of the systematic review process. Results: Results showed that (i) when using the RS/AT approach, less experienced researchers in the subject of the review managed to extract similar data to researchers with more experience; (ii) the amount of data extracted with the approach was higher than using the extraction form; and (iii) the researchers considered helpful the computational support developed. Conclusions:The results of the studies provide insights that the RS/AT approach is feasible in practice, it avoids rework during the data extraction and it leads to a greater amount of extracted data when compared with the use of data extraction form, which is the traditional approach to extract data in the literature review process.