The L16 Experiment ArrayWhat is an 'L16' Experiment Array?An L16 covers all combinations of four parameters at two levels. This is known as a 'full factorial' experiment because the test covers all of the combinations of the values.For this experiment, we varied:
Why did we do it?Even though the coverage of the parameter space is fairly limited, and doing a full factorial test is time consuming, the test gave us the ability to look for interactions between parameters. That is, in addition to showing the average effect of opening the UVG or the DVG, the full factorial test can show if there is any synergistic effect (beyond the individual effects) of changing them together.Also, the redundancy in the test allows estimation of interactions between groups of three factors. Since these are generally non-physical, they can be used to provide a 'free' estimate of the total noise (experimental variation) in the test. This is extremely useful in determining which effects are 'real' compared to normal test-to-test variations.
ResultsThe L16 array has been analyzed using several metrics, and is available on the analysis page of this CD.
For More Information on DOE |