A balanced ANOVA is one where all combinations of factors have the same number of observed results.
Here, we will look at the analysis of a designed experiment. The study is on an automobile filter to reduce pollution. This study looks at the noise properties of the filter. The response of the noise level is in decibels. The factors include vehicle size—1 2 3
, representing small to large, type of filter—1
representing standard and 2
representing octel, and side—1
representing right and 2
representing left.
We will use a balanced ANOVA to study the effect on the results as all treatment conditions have the same number of observations. The terms in the model will be checked for significance using an alpha of 0.05. All main effects and interactions will be included in the model and we will produce residual plots to check the assumptions of using ANOVA.
The data and story behind this example are available from StatLib at http://lib.stat.cmu.edu/DASL/Stories...