4. In an experiment to study the eect of the amount of baking powder in a biscuit dough upon the rise heights of the biscuits, four levels of baking powder were tested and four replicate biscuits were made with each level in a random order. The results are shown in the table below.

(a) What is the experimental unit?

(b) Perform the Analysis of Variance to test the hypothesis of no treatment eect.

(c) Formulate a contrast to test the hypothesis that increase in rise height is proportional to the increase in baking powder in the dough and test this hypothesis.

(d) Estimate the variance of the experimental error 2.

(e) Make a plot of residuals versus predicted values and normal plot of residuals and comment on whether the assumptions of the linear model are justied.

(f) If the dough were made in batches and the four replicate biscuit rise heights in each column (shown in the table above) were all from the same batch, would your answer to

(a) be dierent? How could the data be analyzed if this were the case?

Transcribed Image Text:

25 tsp .5 tsp .75 tsp 1 tsp 11.4 27.8 47.6 61.6 11.0 29.2 47.0 62.4 11.3 26.8 47.3 63.0 9.5 26.0 45.5 63.9