Worked example · Food and Nutrition · Year 11
DCF: Data and computational thinking → Data and information literacy
A secure Year 11 response, annotated so you can see which success criterion each part meets. It is practice only, not NEA work. Answer key: mean line spread 59.3, 42.3, 27.5 and 17.0 mm (15 g is 32.7 mm if the anomaly is wrongly left in); the anomaly is 15 g batch 2, 43 mm; 20 g taste panel means 4.8, 3.8, 3.2, 2.5, 3.7; fall from 10 g to 20 g 59.8%; first choices 15 g: 4, 10 g: 2; rank totals 21, 11, 8, 20. The key idea: the hypothesis is right about thickness, half right about the 20 g sauce, and wrong about which sauce people prefer. Show this after pupils have written their own conclusion.
| Cornflour | Mean of all 3 batches (mm) | Mean without any anomaly (mm) |
|---|---|---|
| 5 g | (58 + 61 + 59) ÷ 3 = 59.3 | 59.3 |
| 10 g | (42 + 44 + 41) ÷ 3 = 42.3 | 42.3 |
| 15 g | (27 + 43 + 28) ÷ 3 = 32.7 | (27 + 28) ÷ 2 = 27.5 |
| 20 g | (16 + 18 + 17) ÷ 3 = 17.0 | 17.0 |
(b) The anomaly is 15 g batch 2, 43 mm. The other two 15 g batches spread 27 and 28 mm, and 43 mm is almost the same as the 10 g sauce. The lab notes explain it: this batch simmered for about 30 seconds instead of 2 minutes. The starch granules did not have time to absorb as much milk and swell fully, so less gelatinisation happened and the sauce was thinner. A controlled variable was not controlled, so the result is not a fair measure of 15 g of cornflour and should be left out. Leaving it in would make the 15 g mean 5.2 mm too high.
(c) 20 g: thickness 29 ÷ 6 = 4.8; smoothness 23 ÷ 6 = 3.8; sweetness 19 ÷ 6 = 3.2; glossiness 15 ÷ 6 = 2.5; starchy taste 22 ÷ 6 = 3.7.
Correct means with working. The anomaly is identified from the pattern, explained with the lab notes and the science, and its effect on the mean is measured. Evidences criterion 1.
(b) As the cornflour increases, the sauce spreads less, so it is more viscous. The mean spread falls from 59.3 mm with 5 g to 17.0 mm with 20 g. From the control recipe (10 g, 42.3 mm) to 20 g the spread falls by 25.3 mm, which is 59.8%. The fall is not the same each time: 17.0 mm from 5 g to 10 g, 14.8 mm from 10 g to 15 g, and 10.5 mm from 15 g to 20 g. Each extra 5 g of cornflour thickens the sauce, but by less each time.
A bar chart suits four separate amounts. The description quotes figures, calculates a percentage change correctly (25.3 ÷ 42.3 × 100), and spots that the pattern levels off, which a quick look at the chart would miss. Evidences criteria 1 and 2.
Key: solid green line = 10 g dashed red line = 20 g
The two biggest differences are starchy taste, 1.5 for 10 g and 3.7 for 20 g (2.2 higher), and thickness, 2.8 and 4.8 (2.0 higher). Glossiness also falls a lot, from 4.3 to 2.5: the 10 g sauce is shiny and the 20 g sauce is dull.
A star diagram shows the whole sensory profile of two samples at once, so the change in shape is easy to see. The comparison uses figures, not just “more” and “less”. Evidences criterion 2.
| Part of the hypothesis | Verdict | Evidence |
|---|---|---|
| More cornflour makes the sauce thicker, so it spreads less. | Supported | Spread falls every time: 59.3, 42.3, 27.5, 17.0 mm. The tasters agree: thickness 1.3, 2.8, 4.2, 4.8. |
| The 20 g sauce will be too thick and lumpy. | Partly supported | Too thick, yes: thickness 4.8 out of 5, “the spoon stands up in it”, and no taster chose it first. Lumpy, no: smoothness was still 3.8 and one taster said “not lumpy, just far too thick to pour”. The problems were a starchy taste (3.7) and a dull look (glossiness 2.5). |
| Tasters will prefer the 10 g sauce. | Not supported | 15 g was preferred: 4 of 6 first choices and the lowest rank total, 8, against 11 for 10 g. “Just right on a crumble.” The 10 g sauce was “a bit thin for a pudding”. |
The science: when the sauce is heated, the starch granules in the cornflour absorb milk, swell and break open. This is gelatinisation, and it traps the liquid so the sauce flows less easily. More cornflour means more starch granules to absorb the same 250 ml of milk, so the sauce becomes more viscous. With 20 g there is so much starch that the sauce becomes a paste instead of a pouring sauce.
Rewritten hypothesis: “As the cornflour in 250 ml of milk increases from 5 g to 20 g, the sauce will become more viscous and spread less, but each extra 5 g will have a smaller effect. Tasters will prefer a sauce with about 15 g for a pudding. At 20 g the sauce will be too thick to pour, with a starchy taste and a dull appearance, but it will not be lumpy.”
Each part of the hypothesis is judged separately against figures and quotations, the thickening is explained with gelatinisation, and the hypothesis is modified to fit every piece of evidence. Evidences criteria 3 and 4.
| Problem | Validity or reliability? | Improvement | |
|---|---|---|---|
| 1 | The 20 g sauce had cooled for about 15 minutes and the 5 g for about 3. Starch sauces thicken as they cool, and cooler food tastes less sweet, so some of the 20 g sauce’s thickness and low sweetness may be caused by temperature, not cornflour. | Validity | Serve every sample at the same temperature, for example 60 °C checked with a food probe, or make each sauce just before it is tasted. |
| 2 | The pots were labelled with the amount of cornflour, and the tasters are friends who may know the hypothesis. They could score what they expect, not what they taste. | Validity | Label samples with random 3-digit codes and do not tell tasters what is being tested. |
| 3 | Every taster tasted in the same order, 5 g first and 20 g last. The last sample is always judged against the others and after the palate is tired. | Validity | Give each taster a different order, so that each sauce is tasted first by some tasters and last by others. |
Also: only 6 tasters, all from one class, is a small panel, which makes the sensory results less reliable. A panel of at least 10 would be better.
Done well: each sauce was made in 3 separate batches, and each result is the mean of 4 readings taken at a fixed 70 °C and 60 seconds. That is why the batches for each sauce agree to within 3 mm, apart from the anomaly, and why the anomaly stands out.
Each problem explains how it could change the results and is correctly labelled as validity or reliability, each fix is specific, and the response credits what Cerys did well. Evidences criterion 5.