Impact of context in visual evaluation of design pastry: Comparison of real and virtual

Abstract

[EN] Virtual reality is becoming an opportunity for multidisciplinary collaboration, including sensory science. The main objective of this work was to compare visual expectations generated by five real cakes evaluated in a laboratory testing booth with their virtual versions in a virtualized sensory booth. Real cakes were designed following the current pastries trends, and virtualization was made using a photogrammetric process. The virtual sensory booth was designed using Unity3D software. The participants were immersed in the 3D virtual environment through a head-mounted display (HMD). Data were analyzed using ANOVA mixed model, internal preference mapping and cluster analysis. The effects of context (real and virtual), the order in test session (crossover design), and socio-demographic factors were studied. The results showed no statistically significant differences within real and virtual studied cakes. These results create new perspectives of the potential of this methodology to be used to rate virtual foods in an immersive environment at the same level as real foods. Regarding the impact of the socio-demographic factors on the acceptance of the cakes, only the effects of gender and sweet-tooth were statistically significant. Males assessed all variables associated with the cakes significantly higher (p < 0.05) than females. In the same way, all variables in all cakes were better evaluated by participants with a self-declared sweet-tooth (p < 0.05). In the internal preference map, participants were segmented into three clusters that could be identified with the different trends in the pastries market. These results may help companies build tailor-made marketing strategies using Immersive technologies to evaluate new food products to satisfy different consumer segments.