Choose Language Hide Translation Bar
Highlighted
jwilliams
Level I

Can I change a Random Block to a Blocking Factor post DOE execution

Hello everyone, I hope you are well! I have run into a problem of my own devising. While designing an RSM DOE, instead of adding a Blocking Factor of Day of Run, I designed a DOE with Random Blocks. This caused me to fall into the rabbit hole of REML. Now that the experiment is complete, with data ready to fit...is there a way to change the Random Block to a Blocking Factor? Or do I need to proceed into the world of REML? I have no experience in scripting, so I would prefer an interactive explanation.

Regards,

Jason

1 ACCEPTED SOLUTION

Accepted Solutions
Highlighted
jwilliams
Level I

Re: Can I change a Random Block to a Blocking Factor post DOE execution

Hello Mark,

Thanks for your prompt reply! Exactly what I needed to know. The DOEs we have executed previously for this client use blocking as a fixed effect, and the client will ask us why this one was fit using REML. Trepidation accompanies the pursuit of anything new, but I will overcome and pursue both strategies so I can compare.

Stay well,

Jason

View solution in original post

2 REPLIES 2
Highlighted

Re: Can I change a Random Block to a Blocking Factor post DOE execution

So you want to treat the Day effect (if any) as a fixed effect, like the other factors?

 

What is your trepidation about using REML to estimate the random block effect and other fixed effects?

 

You can simply change the type of effect in the model, assuming that you plan to use the Model table script that was automatically saved when you clicked Make Table. Right-click the table script and select Edit. Delete the " & Random" attribute after the Day term. You are messing with a script, but you do not have to write a script.

 

 

Learn it once, use it forever!
Highlighted
jwilliams
Level I

Re: Can I change a Random Block to a Blocking Factor post DOE execution

Hello Mark,

Thanks for your prompt reply! Exactly what I needed to know. The DOEs we have executed previously for this client use blocking as a fixed effect, and the client will ask us why this one was fit using REML. Trepidation accompanies the pursuit of anything new, but I will overcome and pursue both strategies so I can compare.

Stay well,

Jason

View solution in original post

Article Labels