An Automated Approach to Recording and Analyzing Design Activities Using a Graphical User Interface

Abstract

Design researchers use protocol studies to understand the cognition of designers and how they think. Although different qualitative and quantitative methods have been employed, analyzing protocol data often requires manual coding, which is tedious, time-consuming, and susceptible to human error. To address these challenges, this paper presents an automated method to record and analyze the designer’s actions. To demonstrate the proposed method, we designed and conducted an experimental study that used a graphical user interface to record design actions while the designer solves a factory layout problem. Then, we used spectral clustering to analyze the data and identify the subproblems that the designer considered. The final design layouts were evaluated using two metrics: travel distance and adjacency score. Our automated method to record and analyze designers’ actions provides a framework to analyze protocol study data sets objectively and quickly.