A paint company wants to compare the abilities of four (4) white house paints to withstand environmental conditions. Eight (8) square houses, each with one side facing exactly north, are available in each of three states: Florida, Michigan, and California (that is, there are 24 houses altogether). Each side of a house is possibly exposed to different types of weather. Also, the houses are different from each other (because of different building materials, different ages, etc.). The company wants to paint each side of each house with a different paint. In addition to comparing the 4 paints the company is also interested in finding our whether differences among the paints vary from state to state. (i) Describe how you would set up the experiment, that is, what error-control design would you use. Explain the reasons for choosing the design. (ii) Give an appropriate linear model for analyzing data from the experiment described in (i). (iii) Outline the ANOVA table associated with the model given in (ii) (giving source of variation, d.f.) and indicate how you would investigate the ques- tions the company is interested in. (iv) Describe how you would perform the analysis in (iii) by using SAS or some other statistical package.