Leveraging the power of eye-tracking for virtual prototype evaluation: a comparison between virtual reality and photorealistic images

• Combining eye-tracking with self-report methods provides valuable insights into product perception and user behavior. • Integrating eye-tracking with self-report questionnaires may enhance the product design process by providing more comprehensive and accurate feedback from the user. • The presentation medium affects both product perception and user observation patterns. • The user’s background (e.g., university of origin) can also influence product assessment. • A joint assessment can mitigate the impact of the viewing medium on user perception, leading to less biased feedback. Most of the information we gather from our environment is obtained from sight, hence, visual evaluation is vital for assessing products. However, designers have traditionally relied on self-report questionnaires for this purpose, which have proven to be insufficient in some cases. Consequently, physiological measures are being employed to gain a deeper understanding of the cognitive and perceptual processes involved in product evaluation, and, thanks to their integration in Virtual Reality (VR) headsets, they have become a powerful tool for virtual prototype assessment. Still, using virtual prototypes raises some concerns, as previous studies have found that the medium can influence product perception. These results rely solely on self-report techniques, highlighting the need to explore the use of ET for product assessment, which is the main objective of this research. We present two case studies where a group of people assessed through two display mediums (CS-1) a set of furniture comprising a general scene using a ranking-type evaluation (i.e., joint assessment) and (CS-2) two armchairs individually using the Semantic Differential technique. Moreover, the dwell time of the Areas of Interest (AOIs) defined was recorded. Primarily, our results showed that, despite VR being sensitive to aesthetic differences between designs of the same product typology, the medium may still influence the perception of specific product attributes —e.g., fragility (p MODERN < 0.001, p TRADITIONAL = 0.002)—, and observation of specific AOIs —e.g., AOI1 (p MODERN = 0.003, p TRADITIONAL < 0.001), AOI9 and AOI10 (p < 0.001). At the same time, no differences were found in the perception of the general scene, whereas dwell time was influenced for AOI1 (p = 0.003), AOI4 (p = 0.006), and AOI5 (