Use of the IRaMuTeQ software to analyze society’s contributions in a regulatory process conducted by the National Health Regulatory Agency

• Autor(a) principal: Carvalho, Thiago Silva
• Data de Publicação: 2020
• Outros Autores: Mota, Daniel Marques, Saab, Flávio
• Tipo de documento: Artigo
• Idioma: por; eng
• Título da fonte: Vigilância Sanitária em Debate
• Texto Completo: https://visaemdebate.incqs.fiocruz.br/index.php/visaemdebate/article/view/1429
• topic: Agência Nacional de Vigilância Sanitária; Participação Social; Análise de Impacto Regulatório; Software IRaMuTeQ; Rotulagem de Alimentos; Brazilian Health Regulatory Agency; Social Participation; Regulatory Impact Assessment; Software IRaMuTeQ; Food Labeling

Introduction: The use of software for analyzing texts arising, for example, from open questions contained in electronic forms has been recommended in the scientific literature. Objective: To demonstrate the use of IRaMuTeQ software, as a support tool for the analysis of social contributions of regulatory process conducted by the Brazilian Health Regulatory Agency (Anvisa). Method: Case study that used a textual corpus composed of the answers of the society about the preliminary report of the Regulatory Impact Analysis of the Nutrition Labeling that was submitted to the Public Subsidy Taking (TPS) by Anvisa. Contributions were organized into two corpus, according to two groups (12 stakeholders and the general public), which underwent the following analyses in IRaMuTeQ: lexicographic analysis to determine the frequency and distribution of active words; descending hierarchical classification (CHD) to categorize active words in classes; and correspondence factor analysis (CFA) to visualize the proximity of words and classes from CHD. Results: 346 contributions were analyzed, of which 12 were part of the textual corpus of the 12 stakeholders. Four words (being, consumer, information and nutritional) of the five most frequent were equal in both textual corpora. CHD analyses of both corpora resulted in three major groups, with a slight difference in the number of lexical classes produced. CFA resulted in three lexical worlds for both textual corpora. Conclusions: The lexical analyses, using IRaMuTeQ, allowed to discriminate the argumentative tendencies as well as to understand the relations between words and classes. The IRaMuTeQ can be considered a useful tool to support the routine analysis of open questions provided in forms submitted to social participation mechanisms promoted by Anvisa. The use of software can make decision-making more agile and reliable, since it allows the public authority to know and consider the contributions of society presented in the participatory process.