Avaliação das soft skills dos gestores de projetos de desenvolvimento de software: uma abordagem multicritério linguística

Abstract

Human skills or behavioral competencies known as soft skills refer to how a person relates and interacts with other people. Behavioral competencies are increasingly important skills for information technology (IT) professionals. Therefore, this study aimed to develop a linguistic multi-criteria model (MMLSS) to assess the most relevant soft skills of software development project managers. Through a systematic review of the literature (2006–2024), twenty-five (25) articles were analyzed, identifying thirty-four (34) soft skills grouped into five dimensions: Effective communication seven (7) soft skills; Leadership and motivation ten (10); Empathy and emotional intelligence six (6); Conflict resolution five (5) and Time management and prioritization six (6). In order to identify and evaluate the most and least relevant soft skills and those that have control indicators in organizations, a case study was conducted with eleven (11) Information Technology specialists, including managers, business owners, development supervisors, and recruiters. The research instrument, developed using a Google Forms questionnaire, was divided into four (4) sections: participant and organization profiles, identification of the most relevant soft skills, indication of the least relevant soft skills, and survey of soft skills with control indicators. The MMLSS uses the 2-tuple linguistic approach and incomplete additive linguistic preference relations (RPLAI), which enable more accurate assessments. Based on the algorithm by Yao and Khalid (2018), the importance weights for each dimension were defined: Effective communication: 0.327; Leadership and motivation: 0.156; Empathy and emotional intelligence: 0.175; Conflict resolution: 0.144; and Time management and prioritization: 0.198. The results obtained from the case study showed that effective communication is the most relevant soft skill for software project managers, according to the proposed model.