Microsoft Office Tutorials and References
In Depth Information
Exhibit 6.9 shows the dialog box to perform the analysis. The Input Range is the
entire table, including labels, and the level of significance,
,is0.05
The results of the ANOVA are shown in Table 6.13. The upper section of the
output, entitled SUMMARY , shows descriptive statistics for the two factors in the
analysis—Groups (A–E) and Products (a–d). Recall that we will be interested only
in the single factor, Products, and have used the blocks to mitigate the extraneous
effects of skill. The section entitled ANOVA provides the statistics we need to either
accept or reject the null hypothesis: there is no difference in the task completions
times of the 4 software products. All that is necessary for us to reject the hypothesis
is for one of the four software products task completion times to be significantly dif-
ferent from the others. Why do we need ANOVA for this determination? Recall we
used the t-Test procedures for comparison of pair-wise differences—two software
products with one compared to another. Of course, there are 6 exhaustive pair-wise
comparisons possible in this problem—a/b, a/c, a/d, b/c, b/d, and c/d. Thus, 6 tests
would be necessary to exhaustively cover all possibilities. It is much easier to use
ANOVA to accomplish the same analysis as the t-Tests, especially as the number of
pairwise comparisons begins to grow large.
α
Table 6.13
ANOVA for analyst example
 
Search JabSto ::




Custom Search