Example of Fully Nested ANOVA

A manufacturing engineer wants to understand the sources of variability in the manufacture of glass jars. The engineer's company manufactures the glass jars at four locations. Four operators measure furnace temperatures in three batches over four shifts from the four locations.

The operators at each plant are different, so the operator factor is nested in the plant factor. While each shift number represents the same part of the workday, the shifts that each operator works at the same plant are different. Thus, shift is nested in operator. Also, the batch of material that the operators use changes each shift. Thus, batch is nested in shift. Because of the nesting pattern, the engineer uses fully nested ANOVA so that the model specification in Minitab is easier.

  1. Open the sample data, FurnaceTemperature.MTW.
  2. Choose Stat > ANOVA > Fully Nested ANOVA.
  3. In Responses, enter Temp.
  4. In Factors, enter Plant-Batch.
  5. Click OK.

Interpret the results

The ANOVA table indicates that the main effects for plant and shift are statistically significant at the 0.05 significance level. The operator effect is not statistically significant at the 0.05 level. The variance component estimates indicate that the variability attributable to batches, shifts, and plants was 52%, 27%, and 18%, respectively, of the total variability.

Analysis of Variance for Temp

SourceDFSSMSFP
Plant3731.5156243.83855.8540.011
Operator12499.812541.65101.3030.248
Shift481534.916731.97742.5780.000
Batch1281588.000012.4062   
Total1914354.2448     

Variance Components

SourceVar Comp.% of TotalStDev
Plant4.21217.592.052
Operator0.8063.370.898
Shift6.52427.242.554
Batch12.40651.803.522
Total23.948  4.894

Expected Mean Squares

1Plant1.00 (4) + 3.00 (3) + 12.00 (2) + 48.00 (1)
2Operator1.00 (4) + 3.00 (3) + 12.00 (2)
3Shift1.00 (4) + 3.00 (3)
4Batch1.00 (4)