Two different plants that produce iced coffee for a particular food company used their in-house sensory panels for evaluation of sweetness. The results are shown in the table below. The QA manager, who is cautious by nature, considers the data to be ordinal rather than interval or ratio.

(a) Using the appropriate non-parametric test, determine if there is a difference in sweetness between the drinks produced at the different plants (-level =0.05). (8)

(b) If the QA manager considered that the data were of interval quality, what would be the appropriate test to use? (1)

(c) Assume that the data are of interval quality and use the test in (b) to determine if there is a difference in sweetness between the drinks produced at the different plants. Compare this result with what you concluded for part (a). (6)

(d) Sometimes non-parametric and parametric tests like those used in parts (a) and (c) give a different answer for the same data set. Why is this so? (2)

Sweetness of dairy drinks from two different manufacturing plants.

PLANT
Eastern	Western
16.8	14.3
12.9	19.9
16.0	19.3
14.9	15.3
16.1	21.3
19.4	19.7
13.7	13.6
18.0	19.8
15.7	21.0
12.7