Practical Application, Investigation and Evaluation Skills
What this domain tests
Whether you can plan and carry out a small practical task or investigation, handle data honestly and evaluate your own work against its aim.
Core rules
- Structure: aim, testable hypothesis, method, results, analysis, conclusion, evaluation.
- Variables: independent (what you change), dependent (what you measure), controlled (held constant). Confounding variables destroy validity.
- Methods include experiment, survey, interview, observation and sensory evaluation - pick the one that actually answers the question.
- Questionnaires: mainly closed questions for analysis, one or two open, no leading or double-barrelled questions, pilot on about five people, avoid jargon.
- Sampling: random, opportunity or stratified; small or self-selected samples cannot be generalised from.
- Reliability means consistency on repeat (repeat trials, standard method). Validity means measuring what you claim (controls, calibrated instruments).
- Sensory tests: nine-point hedonic scale for liking, ranking for preference order, triangle test for detecting a difference (needs a trained panel), star profile for describing attributes scored 0–5.
- Analysis: mean, median, mode and range; percentages; choose the right chart - bar for categories, line for trends, pie for proportions.
- Evaluation: judge against the aim, cite evidence, state limitations and propose realistic improvements.
- Ethics: informed consent, confidentiality, safe practice and food hygiene.
- Costing: ingredient cost per portion plus overheads plus margin.
Worked example
Does wholemeal flour change bread volume? Independent variable: flour type (white versus 100% wholemeal). Dependent variable: volume, measured by rapeseed displacement. Controls: 500 g flour, 7 g yeast, 300 ml water at 30 °C, one hour proving at 25 °C, same tin. Repeat each three times and compare means. Bran in wholemeal cuts gluten strands and absorbs water, so a lower volume is expected - but only a controlled, repeated test supports that claim.
Traps that cost marks
- Calling repeated trials "valid" - repetition gives reliability.
- Using an untrained panel for a triangle test.
- Asking "Don't you think the wholemeal loaf was better?", which is leading.
- Concluding from a single trial or a self-selected sample.