[R] lme4 2 factor factorial model with random factors

Ben Bolker bbolker at gmail.com
Fri Dec 19 15:56:02 CET 2014

Jacob Warren (RIT Student <jjw3952 <at> rit.edu> writes:

> Using lme4 how does one define a 2 factor factorial model with both factors
> being random?
> Specifically I am just trying to recreate the results from Montgomery's
> Design of Experiments book (7th edition), example 13.2. In this example
> there are 2 random factors and I want to include the interaction in the
> model as Montgomery tests for significance in the full model first. I've
> tried several things but cannot recreate the results in R. I would think
> something like what's given below would work, but it does not.
> lmer(y ~ (1 | Parts) + (1 | Operators) + (1 | Parts:Operators) )

 You also posted this on StackOverflow (broken URL to make Gmane
happy, sorry): <http://stats.stackexchange.com/questions/129592/
2-factor-factorial-model-with-random-factors>. While cross-posting
between SO and R mailing lists isn't explicitly forbidden, I think
it's polite to indicate you've done so/cross-link. On SO, I
asked for more information -- not all of us have immediate access to
Montgomery's book (maybe you can provide a Google books link if the
text is accessible that way?).

  SO or r-sig-mixed-models at r-project.org are probably more appropriate
follow-up venues.

  Ben Bolker

More information about the R-help mailing list