[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: Re: st: CI for adjusted mean

Subject   Re: Re: st: CI for adjusted mean
Date   Wed, 24 Oct 2007 09:18:54 -0500

Maarten buis <> asks:

> ... I expected that this code:
> sysuse auto, clear
> recode rep78 1/2 = 3
> anova mpg rep78 wei, continuous(wei)
> adjust wei, se ci
> would give me a mean of mpg adjusted for wei (with standard error and
> confidence interval), but instead it gives me an empty table. Does
> anyone have any thoughts on this one?

The description paragraph for -adjust- (manual or help file) says:

    ...  If by() is not specified, adjust produces results as if
    by() defined one group.  Variables used in the estimation
    command but not included in either the by() variable list or
    the adjust variable list are left at their current values,
    observation by observation.  Here adjust displays the average
    estimated prediction (or the corresponding probability or
    exponentiated prediction), substituting the mean of these
    unspecified variables within each group defined by the
    variables in the by() option.

Since -by()- was not specified, we are dealing with one group
(the whole estimation sample) and since rep78 was not specified
in the -adjust- command, the mean of rep78 over the whole
estimation sample is used in the prediction (see the last part of
the description "... substituting the mean ...").

With your example the mean of rep78 over the estimation sample is
3.57971.  After an -anova- where rep78 is in the model as a
categorical variable, -predict- (and hence -adjust- which uses
-predict-) returns a missing value when asked for the prediction
of the model where the categorical variable is set to a value
that was not one of the levels of that categorical variable.
That is why you end up with an empty table--the -predict- command
returned a missing value for the prediction.

You could instead do things like

    . adjust wei rep=3, se ci
    . adjust wei rep=4, se ci
    . adjust wei rep=5, se ci

or more easily

    . adjust wei, by(rep) se ci

With these cases, the rep78 variable is treated as a categorical
variable (like it should be in this -anova- example).

Ken Higbee
StataCorp     1-800-STATAPC

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index