O professor César França, que atua na LC, aprovou, em conjunto com colaboradores, três artigos na conferência ESEM - Empirical Software Engineering and Measurement (Qualis A2), a ser realizada na Espanha. Os dois artigos abaixo são resultantes do trabalho do grupo de pesquisa HASE, que estuda aspectos humanos em engenharia de software:
-
Monteiro, D.; da Silva, F. Q. B.; D. Valença; Freitas, E. L. S. X.; França, C. (2016). Advantages and Disadvantages of using Shared code from the Developers Perspective: A qualitative study. Proceedings of the 10th ACM/IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM). Context : The code ownership has influence on various aspects of software development, such as code quality, cooperation and team knowledge. There are some studies that seek to understand code ownership, however, studies from the point of view of developers are scarce. Aims: to investigate what are the advantages and disadvantages of practicing shared code ownership from the perspective of the software developers. Methodology: A qualitative study was conducted using a semi-structured interviews in three technology companies with different profiles. We conducted 19 interviews, that were audio recorded and then transcribed. We coded the data using qualitative coding techniques. Results: Considering companies’ context, we have found six advantages and six disadvantages of using shared code ownership. Five proposition were presented. Conclusion: It is noted that the practice of shared code ownership, as described in theory, is more suitable for a more experienced programmers teams, who are able to understand codes without assistance. However, the adaptation of the practice, in which the author is consulted before any change is done, it is necessary for less experienced teams, who feel unsecure to modify the code cause other errors for the project.
-
da Silva, F. Q. B.; França, C.; de Magalhães, C. V. C.; Santos, R. E. S. (2016) Preliminary Findings about the Nature of Work in Software Engineering: An Exploratory Survey. Proceedings of the 10th ACM/IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM). Context: Work Design refers to the different ways in which a given work or task can be designed and performed. The study of work design is important because every decision related to how the work is performed can affect the outcomes of individuals and the effectiveness of teamwork. Goal: To investigate work design characteristics of software engineering work and identify areas for future research. Method: We performed a survey on a sample of 80 software engineers working in 35 different software organizations in Brazil. We measured 21 work design constructs, as well as job burnout, role conflict, role ambiguity, and two constructs related to job rotation. Results: Comparisons of our sample with other sample consisting of a diversity of professions showed that software engineering work has distinct characteristics, but some results require further investigation. Conclusion: We identified relevant characteristics of software engineering work and areas for further research. In particular, longitudinal studies are needed to address the temporal variations impossible to identify in cross sectional studies.
Ao todo, o grupo de pesquisa teve 6 artigos aprovados na conferência, 3 completos e 2 resumidos. O grupo de pesquisa também conta com a participação do professor Cleviton Monteiro (do DEINFO), que também foi co-autor em alguns destes artigos.
O terceiro artigo aprovado pelo professor César França foi em conjunto com Bruno Falcão (aluno do Mestrado Profissional em Engenharia de Sofware do CESAR) e consultores do CESAR:
-
França, C.; Peixoto, E.; Falcão, B.; Monteiro, C. (2016) The Obscure Process of Innovation Assessment: A Report of an Industrial Survey. Proceedings of the 10th ACM/IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM) Context - Software companies should track innovation as rigorously as core business operations. For that, the assessment of innovation projects is a critical process, in particular to make their innovation initiatives funded. Objective - In this article, we aim to evaluate the need for more practical measurement tools, by checking the agreement of very experienced analysts, from the industry, about the innovation degree of four actual software projects. Method - We conducted a survey with eight business analysts, using a combination of the Three Horizons Model and the Gartner's Hyper Cycle for emerging technologies as a frame of reference. Results - In general, the level of agreement about the innovation degree in the projects was very low. Looking at the cases in isolation, it is possible to suggest reasons for the low level of agreement between the evaluators. Conclusions - Our data support the fact that innovation is an activity difficult to characterize and even more difficult to measure, and the need for practices to achieve better intersubjective agreement for innovation assessment became evident in this work.