Randomized Block Design Refer to the cell phone study in Example 11.9, in which the effect of usage level on cost (in dollars) was studied for four different companies. 

1. Enter the data into columns A-E of an Excel spreadsheet, using column A for usage labels and row 1 for company labels, just as shown in the table above. 2. Use Data Data Analysis Anova: Two-Factor Without Replication to generate the dialogue box in Figure 11.15(a). Highlight or type the Input Range (the data in the first five columns) into the first box and check "Labels" if neces- sary. Change the significance level if needed, and click OK. You can adjust the output, possibly changing the labels "Rows" and "Columns" to "Usage" and "Company," as shown in Figure 11.15(b). 3. The observed value of the test statistic for treatments (companies) is F = 1.84 with p-value = 0.240 indicating that there is no significant difference among the four companies. The test for blocks (usage) is highly significant, with p-value = 0.000.

Company Usage Level A B C D Low 27 24 31 23 Middle 68 76 65 67 High 308 326 312 300