Learning Agile Estimation in Diverse Student Teams by Playing Planning Poker with the Humanoid Robot NAO. Results from Two Pilot Studies in Higher Education

Abstract

Agile estimation is performed by teams to predict relative effort needed to finish project tasks. Estimating in diverse teams may be challenging due to different subjective perspectives. Planning poker is a game-based technique applied by agile teams to empower all team members to jointly estimate and reach a consensus about the predicted effort. This paper presents results from two studies on learning agile estimation in which student teams were supported by the humanoid robot NAO acting as facilitator of the game. The paper describes the design of the robot-assisted planning poker simulation, the programming of the application, and the evaluation results from two pilot studies with 29 bachelor and master students in different study programs and with different cultural backgrounds. The evaluation aimed to investigate students' perceptions of the robotic facilitator, self-assessment of learning outcomes related to agile estimation, and possible effects of different cultural dimensions on the perceptions of the robot-assisted simulations and the learning outcomes. The results show that both bachelor and master students, independent of their cultural background, perceived the NAO robot as a trustworthy, friendly and likeable facilitator of the planning poker game. Students with a European background rated the possibility of befriending NAO slightly higher compared to non-European students. Students also reported that playing planning poker in teams with support of the robot helped them to understand agile estimation. The participants recommended to use the simulation game “Planning poker with NAO" in future classes on project management and estimation.