Sensory evaluation is a scientific discipline that uses human senses — sight, smell, taste, touch, and hearing — to measure, analyze, and interpret the sensory properties of foods. It plays a critical role in product development, quality control, shelf-life studies, and consumer acceptance testing.
Discrimination Tests
Discrimination tests determine whether a perceptible difference exists between two or more products. The triangle test presents panelists with three samples, two of which are identical, and asks them to identify the odd sample. The duo-trio test provides a reference sample followed by two coded samples, asking which matches the reference. Paired comparison tests ask panelists to evaluate which sample has more of a specific attribute, such as sweetness or bitterness. Discrimination tests are forced-choice (panelists must select an answer even if uncertain) and are analyzed using binomial probability tables.
Descriptive Analysis
Descriptive analysis provides quantitative measurement of all sensory attributes of a product. Quantitative Descriptive Analysis (QDA) uses a trained panel to rate the intensity of attributes on unstructured line scales, with results displayed on a spider plot. The Spectrum method uses reference standards to anchor intensity scales, allowing results to be compared across studies. Both methods require 8–16 trained panelists and multiple replicate evaluations. Panelists develop a common lexicon of descriptors through extensive training, covering appearance, aroma, flavor, texture, and aftertaste.
Hedonic Testing
Hedonic (affective) tests measure consumer liking and preference, typically using large panels of untrained consumers (100–300 individuals) who represent the target population. The 9-point hedonic scale ranges from “dislike extremely” to “like extremely”. Acceptability tests determine whether a product is acceptable to consumers, often using a yes/no or just-about-right (JAR) scale. These tests are essential for assessing market potential and guiding product optimization.
Panelist Selection and Training
The success of sensory evaluation depends on panelist quality. For discrimination and descriptive tests, panelists are screened for sensory acuity (ability to detect basic tastes, identify common odors, and rank intensities), motivation, communication skills, and availability. Training for descriptive panels typically requires 40–100 hours over 8–12 weeks, during which panelists learn to identify, define, and quantify sensory attributes.
Tasting Booths and Environmental Control
Sensory testing should be conducted in a controlled environment to minimize bias. Individual tasting booths, illuminated with standardized lighting (red light to mask color differences when needed), are isolated to prevent panelist interaction. Positive air pressure and activated carbon filtration prevent odor contamination. Samples are presented in randomized, blinded containers with three-digit codes. Data collection uses computerized systems or paper ballots, depending on the facility.
Statistical Analysis
Statistical methods are integral to sensory data interpretation. Analysis of variance (ANOVA) is used to determine whether products differ significantly across attributes. Principal component analysis (PCA) reduces multivariate sensory data to identify relationships between products and attributes. Consumer data may be analyzed by cluster analysis to segment markets, and by internal or external preference mapping to relate consumer liking to descriptive data. Sensory properties are closely linked to food structure, which can be characterized by rheology and texture analysis. The Maillard reaction and lipid oxidation contribute to flavor development assessed by sensory panels.
