How Can I konw if the aliasing exists in the Alias Matrix?

[#1]

How Can I konw if the aliasing exists in the Alias Matrixthe involving two-way interactions and main effects?

For example, in this Alias Matrix, what does 1 and 0 mean?

[#2]

The values in this table are the correlations between the parameter estimates. A 0 means that they are estimated independently. A 1 means that they are confounded. (Each of the parameters is an alias for the same column used to estimate the effect.) A value between 0 and 1 indicates correlation that does not prevent estimation but it will inflate the variance of the estimates, and reduce power.

See Help > Books > Design of Experiments. The information in the Design Evaluation outlines is thoroughly explaned and demonstrated.

[#3]

Thanks for the information.

From the Screening Design, how do I know if there is aliasing?

[#4]

I think you will need to be more specific.

From the menus, are you choosing DOE > Classical > Two Level Screening > Screening Design?

If so, after specifying your factors you are given a choice of choosing a design from a fractional factorial design or to construct a main effects screening design.

If you choose a fractional factorial design, a catalog of designs will appear that provides the resolution. That is usually sufficient, but if not, there is an item in the report window that is labelled Aliasing of Effects. That will show all aliasing up to order 2. If you want higher, click the red triangle at the bottom to choose Show Confounding Pattern to generate to whatever order you wish.

If you choose a main effects screening design, there is an Alias Matrix output in the report which is the correlation matrix that Mark described.