Practical guidance · QualiFood
Which method do I need?
Start with the decision you need to make. This guide distinguishes among 100% inspection, acceptance sampling, statistical process control, design of experiments and risk-based inspection prioritisation.
STEP 1
What is your main question?
Choose the option that best describes the immediate decision. If you have more than one question, repeat the pathway for each one.
QUICK COMPARISON
Five approaches, five decisions
| Method | Decision unit | Core question | When to use it | Do not confuse it with |
|---|---|---|---|---|
| 100% inspection | Each unit | Does this unit comply? | In-line or final inspection | An absolute guarantee: measurement errors can still occur |
| Acceptance sampling | Lot | Accept or reject? | Receipt or release | Process control and improvement |
| SPC | Process over time | Is it stable and predictable? | During operation | Comparison with specifications alone |
| DOE | Experimental space | What causes the change? | Development and improvement | Changing one factor at a time |
| Risk-based prioritisation | Establishment or activity | Where should I inspect first? | Control planning | A definitive compliance assessment |
FOOD SCENARIOS
Test the logic with common situations
Packaging receipt
Decide whether a lot meets the maximum allowable number of sealing defects.
Fill weight
Detect shifts in the mean during production.
Biscuit texture
Study temperature, time and formulation, including their interactions.
Official inspection programme
Assign frequencies according to inherent risk and historical performance.
Guidance, not automated decision-making. The final selection should consider the objective, data design, measurement capability, error risks, legal requirements and professional judgement. Two methods may complement each other within the same system.