From: <arthur.gilmour_at_DPI.NSW.GOV.AU>
Date: Wed, 22 Oct 2008 08:00:51 +1100

Dear Anne,

> I am curious about the use of the MVINCLUDE option in ASReml, especially
in univariate analyses, and would like to clarify when/why it can be
> My understanding is that it can be useful for multivariate analyses when
the design matrix has a missing value, in which case it is set to zero,
which can only work for covariates, which have previously been
mean-centered? Is that correct?
** If the covariate value is missing, you must either ignore nthe data
record, or impute some value to replace the missing value.
The MVINCLUDE qualifier imputes a zero. If the data is centred, this is
reasonable. You might of course want to impute a value
other than the simple mean which could be done using the !NA
transformation, or by 'CENTERING' to the required value.

> If applied to a simple model with one categorical fixed effect (e.g.
sex) and one random effect, such as x ~ mu sex !r animal, am I right to
assume that in the case of missing values for ?sex? the MVREMOVE option
should be always avoided because it would then assign all missing sexes as
males, or all as females?
** MVREMOVE discards the RECORD containing the with the unknown sex,
rather than use the record.
MVINCLUDE may assign missing to one sex or the other, or may define a
third 'sex' depending on whats in the model.
Consider the DESIGN matrix
mu Male Female
1 1 0 for a MALE
1 0 1 for a female
1 0 0 for an unknown.

Now the usual model singularity (mu = Male + Female) is broken so ASReml
will fit 2 DF for SEX and 1 for mu,
that is, the missing sexes become a third sex class.

Using MVREMOVE sometimes messes up the record pattern in multivariate
analyses and should be used with care in that context.

Thanks in advance for any answer,
May Jesus Christ continue to be gracious to you,

Arthur Gilmour, His servant .
Mixed model regression mapping for QTL detection in experimental crosses.
Computational Statistics and Data Analysis 51:3749-3764 at

Personal website:

Skype: arthur.gilmour,
Principal Research Scientist (Biometrics)
NSW Department of Primary Industries
Orange Agricultural Institute, Forest Rd, ORANGE, 2800, AUSTRALIA
fax: 02 6391 3899; 02 6391 3922 Australia +61
telephone work: 02 6391 3815; home: 02 6364 3288; mobile: 0438 251 426

ASREML 2 is available from
Register on the ASReml forum at
ASReml-L Archives are at

The ASReml3 alpha download registration page is:
There are asreml3alpha forums at:

Proposed travel:
Coffs Harbour WS 14-17 October (Forestry)
PMADs Wollongong 21-24 Oct
Received on Fri Oct 22 2008 - 08:00:51 EST

This webpage is part of the ASReml-l discussion list archives 2004-2010. More information on ASReml can be found at the VSN website. This discussion list is now deprecated - please use the VSN forum for discussion on ASReml. (These online archives were generated using the hypermail package.)