Principal Component Analysis of System Usability Scale for Its Application in Automotive In-Vehicle Information System Development

The System Usability Scale (SUS) is used across industries, to evaluate a product’s ease of use. As the automotive industry increases its digital footprint, the SUS has found its application as a simple and reliable assessment of various in-vehicle human machine interfaces. These evaluations cover a broad scope and it is important to design studies with participant fatigue, study time, and study cost in mind. Reducing the number of items in the SUS questionnaire could save researchers time and resources. The SUS is a ten-item questionnaire that can measure usability and learnability, depending on the system. These ten questions are highly correlated to each other suggesting the SUS score can be determined with fewer items. Thus, the focus of this paper is two-fold: using principal component analysis (PCA) to determine the dimensionality of SUS and using this finding to reduce variables and build a regression equation for SUS scores for in-vehicle human machine interfaces. Data from 42 systems were used for this evaluation. Building a prediction model using items 3 (I thought the system was easy to use), 9 (I felt very confident using the system) and 10 (I needed to learn a lot of things before I could get going with the system) was determined to be the best way to reduce the questionnaire while preserving the learnability dimension that occurs in measuring in-vehicle systems. Further studies should be done on how a reduced questionnaire impacts the assumptions built into SUS protocol.