Table 4:
Regression output from the full factorial analysis. A regression is performed to explore the effects of having each variable combination in the model. While weight and dosage are significant, the noise variables are not. This indicates that they should be dropped from the model.
| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 1.15 | 0.01 | 97.36 | 9.22E-19 |
| dosage | 1.15 | 0.01 | 97.24 | 9.36E-19 |
| weight | 0.72 | 0.01 | 61.37 | 2.32E-16 |
| dosage:weight | 0.73 | 0.01 | 61.82 | 2.13E-16 |