Introduction
The UK’s Committee on Toxicity of Chemicals in Food launches an investigation into leaching of a commonly used chemical known as Bisphenol A (BPA) [IUPAC name 4,4′-(propane-2,2-diyl)diphenol]. The chemical is widely used in the manufacture of inert coatings of many types of food packaging. In concentrated form, there is evidence that BPA has hormone-like properties e.g. mimicking oestrogen. This has raised concerns about its presence in consumer products.
This investigation focuses on two different fruit juice carton packaging techniques (A and B), both using cardboard coated with processes involving BPA. In order to test whether the fruit juice quality was adversely affected by the packaging process, an experiment was undertaken comparing the level of BPA from the same batch of fruit juice which was randomly allocated to cartons made using the two techniques. The experimental hypothesis was that BPA was leaching out of the packaging material and entering the fruit juice and that there would be a difference between the two packaging processes with respect to the BPA level.
In total 67 cartons of fruit juice were prepared, 36 using method A and 31 using method B. After being kept in identical conditions (same time, same size/shaped carton, the same volume of fruit juice), the cartons were opened and the fruit juice analysed for BPA using High-Performance Liquid Chromatography (HPLC) via the agency of an accredited analytical chemistry laboratory.
The problem
The Committee needs to know what is the evidence, based on the level of BPA in the fruit juice, whether there is a significant difference between the two carton packaging processes. It is your task to report on this question. You have been presented with data from the 67 analysed packages of fruit juice. Three variables are presented below (Table 1): the sample identifier number, the BPA concentration (in µg/L), and the type of packaging process/method.
TABLE 1
BPA concentrations (µg/L) in fruit juice following packaging by two methods
ID [BPA] µg/L Method
1 9.49 A
2 5.50 A
3 5.59 A
4 3.47 A
5 5.87 A
6 6.05 A
7 9.17 A
8 1.95 A
9 2.79 A
10 8.21 A
11 5.33 A
12 4.43 A
13 4.84 A
14 3.94 A
15 2.22 A
16 4.41 A
17 5.26 A
18 5.55 A
19 4.39 A
20 4.47 A
21 6.39 A
22 4.45 A
23 6.19 A
24 4.09 A
25 8.39 A
26 5.89 A
27 6.82 A
28 5.44 A
29 7.47 A
30 1.98 A
31 6.58 A
32 4.15 A
33 5.34 A
34 7.63 A
35 10.17 A
36 4.77 A
37 3.27 B
38 10.24 B
39 3.89 B
40 8.49 B
41 5.91 B
42 6.71 B
43 5.54 B
44 7.88 B
45 3.62 B
46 7.75 B
47 5.46 B
48 6.94 B
49 7.36 B
50 6.41 B
51 7.60 B
52 4.73 B
53 4.95 B
54 6.33 B
55 9.41 B
56 3.79 B
57 7.64 B
58 7.72 B
59 7.78 B
60 6.30 B
61 7.81 B
62 7.50 B
63 5.61 B
64 3.95 B
65 9.72 B
66 7.14 B
67 6.29 B
Your task
You are required to produce a coherent, concise, professionally presented, and structured report which informs the Committee whether or not there is genuine evidence that the two packaging processes differ from each other in terms of the extent of BPA leaching. In concluding the report you cannot assume that the Committee understands statistical methods in detail and you will therefore have to summarise your findings in plain non-technical language. The total length of the report must not exceed 1250 words (word count excludes cover page, table and figure contents/captions, in-text citations, and references – appendices are not permitted).
You are expected to:
- introduce the type of question you are facing including hypothesis testing
- produce appropriate descriptive and/or graphical (fully formatted) summaries of the data
- justify your choice of statistical test
- undertake the correct statistical tests
- fully interpret statistical test output
- consider the design of the study
- arrive at brief and clear conclusions that directly address the problem
