“It's not bad, but it's not chicken”: Sensorial drivers of liking and their correlates for plant-based chicken analogues in Sweden

Development of satisfying plant based (PB) meat analogues could be expedited through better understanding of the sensorial drivers of hedonic (dis)liking. To address this, here six commercially available PB chicken analogues and a chicken breast reference are evaluated. Consumers (N = 105) report hedonic liking (appearance, aroma, taste/flavour, texture, and overall), sensorially profile the samples with rate-all-that-apply (RATA) and just-about-right (JAR) scales and describe what they like and dislike about the tasted samples in their own words (free-text responses). The samples are also characterized instrumentally: microscopy to visualize the microstructure and cyclic compression testing to quantify key mechanical properties (maximum true stress, Young's modulus, and compressive energy). The sensory space is explored with principal component analysis and penalty analysis; associations between the modalities of liking and how these may determine overall liking is assessed using Bayesian network analysis; and multiple regression with Bayesian inference is used to contrast liking across samples and identify the drivers of (dis)liking within each hedonic modality. Chicken is preferred across all hedonic modalities, while samples made of mycoprotein are liked most of the PB samples. Drivers of liking are detected for all modalities except appearance. The drivers of texture liking include higher perceived tenderness and juiciness, the latter of which is also inversely correlated with Young's modulus on the first compression cycle. The free-text responses suggest that consumers readily recognized the chicken sample. That sensory responses reflect both the intrinsic properties of products (bottom-up) and cognitive (top-down) aspects, including familiarity, is thus discussed.