MICOBISE - Mitigating cognitive biases in software engineering

The MICOBISE project will investigate software engineers' cognitive biases, which are recurring errors in thinking or patterns of bad judgment, and introduce de-biasing practices as mitigation mechanisms. Software development is a socio-technical activity that extends beyond technical matters, involving stakeholder interaction and decision-making under uncertainty. These are prone to the ill effects of cognitive biases, leading to significant overruns in the cost and schedule of software engineering projects and poor-quality outcomes in terms of software-based products and services, negatively affecting end-user experience and their quality of life.

This project aims to develop an evidence-based understanding of cognitive biases in the software development context, incorporating de-biasing techniques to mitigate the harmful effects of cognitive biases in critical decision-making at multiple levels through the software development lifecycle and improve the quality of technical and process decisions.

This project will use a multi-method approach comprising empirical research methods and data collection activities to achieve this goal, specifically literature surveys and interviews for data collection, grounded theory for qualitative data analysis and theory generation, controlled experiments to evaluate generated theories, and statistical data analysis methods for quantitative experiment data analysis.

This project expects to significantly advance our understanding of the socio-technical aspects of software engineering concerning cognitive biases. Expected outcomes are theories on how biases affect software engineers and empirically validated practices to counter these biases. These should result in a paradigm shift in how practitioners develop software and how researchers study software development. Considering the pervasiveness of software across all sectors of the economy and our daily lives, these should provide significant benefits, including reduced development costs, increased productivity, and improved end-user experience with higher-quality software products.