Gage R&R Study (Crossed)

ANOVA Method
Two-Way ANOVA Table with Interaction

  

Two-way ANOVA (analysis of variance) tells you if different levels of two factors result in different average values of a response variable.

The two-way ANOVA table lists the following sources of variability:

·    Part, which represents the variability in measurements across different parts.

·    Operator, which represents the variability in measurements across different operators.

·    Operator*Part, which represents the variability in measurements across different combinations of operator and part. If the Operator*Part term has a p-value greater than 0.05, the ANOVA will be re-run without the interaction term.

·    Error or Repeatability, which represents the variability in measurements that cannot be attributed to parts, operators, or the operator by part interaction.

You want to see if the p-value associated with each source is less than the chosen level of significance. This would indicate that the source contributes significantly to the measurement variability.

Example Output

Source           DF       SS        MS        F      P

Part              9  2.05871  0.228745  39.7178  0.000

Operator          2  0.04800  0.024000   4.1672  0.033

Part * Operator  18  0.10367  0.005759   4.4588  0.000

Repeatability    30  0.03875  0.001292

Total            59  2.24913

 

 

α to remove interaction term = 0.05

Interpretation

For the parts data, the p-value for Operator (P = 0.033) shows you that the average value of thickness is significantly different among the three different operators at the a = 0.05 level of significance.

Here, the Operator*Part term is included because P = 0.000 is less than 0.05. The interaction between operator and part will be taken into account in the Gage R&R.