Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

- JMP User Community
- :
- Discussions
- :
- Blocking in JMP

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Blocking in JMP

Oct 28, 2018 4:42 PM
(5705 views)

I have a set of data where I want to perform a three-way ANOVA in JMP. My experiment has three factors (A, B, C) with response (Y). Each factor has two levels (-1, 1) and two replicates are run. I used "Analyze" -> "Fit Model" and select "Full Factorial" to obtain my three-factor ANOVA. However, I am trying to perform a second test where I block the replicates. How would I go about doing this? Would I used the same "Analyze" -> "Fit Model" methods?

Thank you in advance!

4 REPLIES 4

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Blocking in JMP

Documentation on how to define and run Block Designs can be found in the "Fitting Linear Models" and the "DOE Guide". Both of these books are available under the Help Pull Down Menu

Help==>Books

Help==>Books

Jim

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Blocking in JMP

The blocking for a fixed or random effect is usually included in the design. How was the design blocked?

If you used JMP to design your experiment, for example Custom Design, then the blocking effect would have been properly represented with a term in the model that was is automatically included in your linear regression analysis. Which type of design did you use?

In fact, even f you deem that the blocking effect is null then you should not remove this term in the model, as blocking introduces a restriction on the randomization of the experiment. The blocking of runs is always part of the analysis.

How many runs are in your design with replicates?

Learn it once, use it forever!

Highlighted
##

I am trying to block for a random effect. I have a set of data (Factors A, B, C and Response Y) - each factor has two levels. 2 replicates were run and I have 16 rows of obersvations (that include the replicates). I didn't use the DOE, instead i justed used Fit By Model to run the three factor ANOVA with full factorial to determine significance taking into account replicates. I'm unsure how to block replicates - I see on Fit Y by X you can block a treatment, but I want to block replicates which is confusing me. Sorry if it is not clear, thanks in advance for your help!

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Blocking in JMP

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Blocking in JMP

Created:
Oct 30, 2018 3:54 AM
| Last Modified: Jan 30, 2020 11:54 AM
(5646 views)
| Posted in reply to message from kwyl22 10-29-2018

I am not sure that you need to determine a "block effect." The replicates are used to estimate the error as residuals. They form the Sum of Squared Errors in the Analysis of Variance table.

Introducing blocking in the design associates a specific change in an external variable (not a factor) with a group (block) of runs. Examples might be different equipment, different batches or lots, different sites, different days, and so on. What is the nature of your blocks?

You must create a new data column for the blocks and enter the levels for each run. Then you can add this column to the list of effects in the Fit Model dialog box. Note that (A) you cannot cross this variable with other effects and (B) you should not remove this effect even if you deem it insignificant.

Learn it once, use it forever!

Article Labels

There are no labels assigned to this post.