Quality Assurance and Sensory Evaluation
This domain separates quality control from quality assurance, matches the correct sensory test to the question asked, and tests statistical interpretation of sensory results. Expect questions distinguishing difference, descriptive and affective tests, panel selection, sampling, calibration, traceability and the link between QA and HACCP.
QA, QC and specifications
Quality control inspects and tests product against specifications (moisture, pH, aw, microbial counts, weight, colour). Quality assurance is the wider system of procedures, audits, supplier approval, training and documentation that prevents defects. Specifications must be measurable, with defined methods and acceptance limits.
Choosing the sensory test
- Difference tests: triangle (three coded samples, two identical), duo–trio (reference plus two), paired comparison, ABX. Ask: is there a detectable difference? Need trained or at least screened panellists; typically 20–30.
- Descriptive tests: quantitative descriptive analysis, profiling, ranking for intensity. Trained panels of about 8–12; ask: how does it differ?
- Affective tests: 9-point hedonic scale, paired preference, ranking for liking. Untrained consumers, 60+ preferred; ask: which is liked?
Worked example
A triangle test uses 12 panellists and 7 correctly identify the odd sample. The critical number for significance at α = 0.05 with n = 12 is 8 correct. Since 7 < 8, the difference is not statistically significant - you cannot claim the reformulated product differs. Reasoning: in a triangle test the chance of guessing correctly is 1/3, so the critical values are higher than a simple 50 % majority; you must compare against the table for that n and α, not against half the panel.
Supporting systems
- Sampling: representative sampling plans (e.g. AQL-based), correct sample size, aseptic handling.
- Calibration: instruments checked against standards; uncalibrated pH meters invalidate results.
- Documentation and traceability: batch codes, records, recall procedures.
- Standards: GMP, SSOP, HACCP, ISO 9001 (quality management), ISO 22000 (food safety management).
Panel and design controls
Randomise sample order and use three-digit codes to prevent carry-over and position bias; use blinding and controlled lighting; rinse between samples; do not use a hedonic consumer panel to detect a subtle formulation difference, nor a trained panel to decide market acceptability.
Traps
Confusing QA with QC; using the wrong panel type for the test's purpose; interpreting triangle results by simple majority; ignoring serving order, temperature and coding bias; and testing only one batch, which says nothing about process consistency.