This add-in conducts a hypothesis test for evidence of quadratic curvature in a factorial DOE with center points. It can help you decide if axial points should be added to the design to estimate the curvature of individual factors. Creation of this add-in was inspired by this JMPer Cable post by my colleague @MikeD_Anderson, which is certainly worth a look if this add-in is relevant to you.

The add-in implements the pure quadratic curvature test from Ch. 6 of Montgomery's *Design and Analysis of Experiments* (8e consulted here)*, *specifically the t-test version detailed in Section 6-8 of the supplemental materials. This tests the mean of the factorial points against the mean of the center points while using the variance of the center points as the error estimate. The use of center point variance to estimate error distinguishes this test from the pooled and unequal variances t-tests available in Fit Y by X.

If you teach DOE with Montgomery's *Design and Analysis of Experiments*, this add-in could be a useful tool for your course when covering center points and curvature.

See the attached documentation (*Pure Quadratic Curvature Test Addin.pdf*) for detailed instructions on using the add-in.

Minimum JMP Version: JMP 16

In addition to the add-in and documentation files, you'll find two sample data sets:

*Montgomery Example 6-7.jmp* is the data from the relevant example provided in Montgomery's book.
*Reactor with Center Points.jmp* is a version of the *Reactor 32 Runs.jmp *data from the Sample Data Library that has been augmented with 5 center points.