Meaning
Statistical models evaluate variance components when factors are organized in a hierarchical structure where each level of a factor appears within only one level of another factor. Using a nested anova allows engineers to distinguish between different sources of error, such as the difference between several production lots versus the difference between individual units within those lots. This distinction is essential for identifying the root cause of a quality issue.
Variance Source
The analysis breaks down the total observed fluctuation into specific buckets. When performing a nested anova, the model identifies whether the bulk of the variation comes from the testing equipment, the assembly operator or the raw material supplier. This prevents the team from wasting resources on the wrong part of the process.
Experimental Design
Proper data collection requires a clear hierarchy where subgroups are clearly defined.
Data Interpretation
Results are typically presented as a table showing the mean squares and the significance level for each nested level. A high variance at the top level of the nested anova suggests a problem with the overall process environment or supplier quality. If the variance is higher at the bottom level, the issue is likely related to random noise or local instability in the manufacturing cell.
