Video 6. How to Construct ANOVA Table from the Effect Estimates from the Factorial Design of Experiments DOE.

ANOVA table can be constructed using the calculated main and the interaction effects. Details will be discussed in the “Module 5: 2K factorial design of experiments.” For this particular example with only two independent variables (factors) having two levels for each factor without any replications, the sum of square is simply calculated as the square of the effects. For an example, the sum of square for the temperature variable (factor) is the square of the temperature effect of an increase in comfort level of 6, which is 36 (6²). No experimental error can be calculated without replication. Therefore, the experimental error is zero (0) for this example. Therefore, the F-statistics and the p-value cannot be calculated for this example. The ANOVA table can be constructed as in Table 2.

*need replications (experimental error) to calculate these values

Table 2 ANOVA table from the calculated main and interaction effects