Re: [R] lsmeans

From: John Fox <>
Date: Sun, 08 Jun 2008 15:26:12 -0400

Dear Hadley,

Unfortunately, the term "marginal" gets used in two quite different ways, and Searle's "population marginal means" would, I believe, be more clearly called "population conditional means" or "population partial means." This is more or less alternative terminology for "least-squares means" (to which Searle rightly objects).


John Fox, Professor
Department of Sociology
McMaster University
Hamilton, Ontario, Canada

> -----Original Message-----
> From: hadley wickham []
> Sent: June-08-08 2:52 PM
> To: Douglas Bates
> Cc: John Fox; Dieter Menne;
> Subject: Re: [R] lsmeans
> On Sun, Jun 8, 2008 at 12:58 PM, Douglas Bates <>
> > On 6/7/08, John Fox <> wrote:
> >> Dear Dieter,
> >>
> >> I don't know whether I qualify as a "master," but here's my brief take
> >> the subject: First, I dislike the term "least-squares means," which
> to
> >> me like nonsense. Second, what I prefer to call "effect displays" are
> just
> >> judiciously chosen regions of the response surface of a model, meant
> >> clarify effects in complex models. For example, a two-way interaction
> >> displayed by absorbing the constant and main-effect terms in the
> interaction
> >> (more generally, absorbing terms marginal to a particular term) and
> setting
> >> other terms to typical values. A table or graph of the resulting
> >> values is, I would argue, easier to grasp than the coefficients, the
> >> interpretation of which can entail complicated mental arithmetic.
> >
> > I like that explanation, John.
> >
> > As I'm sure you are aware, the key phrase in what you wrote is
> > "setting other terms to typical values". That is, these are
> > conditional cell means, yet they are almost universally misunderstood
> > - even by statisticians who should know better - to be marginal cell
> > means. A more subtle aspect of that phrase is the interpretation of
> > "typical". The user is not required to specify these typical values -
> > they are calculated from the observed data.
> >
> How does Searle's "population marginal means" fit in to this? The
> paper describes a PMM as "expected value of an observed marginal mean
> as if there were one observation in every cell." - which was what I
> thought happened in the effects display. Is this a subtly on the
> definition of typical, or is that PMM's are only described for pure
> ANOVA's (i.e. no continuous variables in model)?
> Hadley
> --
> mailing list PLEASE do read the posting guide and provide commented, minimal, self-contained, reproducible code. Received on Sun 08 Jun 2008 - 19:57:03 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Sun 08 Jun 2008 - 21:31:17 GMT.

Mailing list information is available at Please read the posting guide before posting to the list.

list of date sections of archive