{smcl}
{* 31aug2006}{...}
{hline}
help for {hi:devcon}
{hline}
{title:Deviation contrast transformation for estimation results}
{p 8 14 2}{cmd:devcon} [ {cmd:,} {cmdab:g:roups:(}{it:glist}{cmd:)}
{cmdab:eq:uations:(}{it:numlist}{cmd:)}
{cmd:check}[{cmd:(}{it:#}{cmd:)}] {cmdab:non:oise} {cmdab:l:evel:(}{it:#}{cmd:)} ]
where {it:glist} is
{p 8 8 2}{it:varlist1} [{cmd:(}{it:varname1}{cmd:)}] [{cmd:,} {it:varlist2}
[{cmd:(}{it:varname2}{cmd:)}] [{cmd:,} {it:...}] ]
{title:Description}
{p 4 4 2} A categorical regressor is usually included in a regression model
using a set of 0/1 dummies differentiating the effects of the
separate categories of the variable. A coefficient associated with such a
dummy variable reflects the expected outcome difference between the
represented category and some reference category. Since one of the
categories serves as the reference category, only {it:k}-1 dummy
variables are used for a {it:k}-category variable.
{p 4 4 2}{cmd:devcon} may be used to transform the coefficients of such 0/1
dummy variables so that they reflect deviations from the "grand mean" (in
other words, the modified coefficients will sum up to zero over all categories)
rather than deviations from the reference category. The transformed
coefficients are equivalent to those obtained by using the so called
"effects coding" for the dummy variables (see the {cmd:e} prefix in
{cmd:xi3} or the {cmd:dev()} contrast in {cmd:desmat}; both packages are available
from the SSC Archive). However, {cmd:devcon} reports
coefficients for {it:all} categories (including the category that was
used as the reference category in the original model) and modifies the
model's constant accordingly (with the effects coding, the coefficient of
one of the categories is "hidden" in the constant). Furthermore, the coding
of the underlying dummy variables is still 0/1 with {cmd:devcon}.
{p 4 4 2}The deviation contrast transformation is applied to the last
(i.e. currently active) estimates. Use the {cmd:groups()} option to define
the group(s) of dummy variables. {cmd:devcon} specified without the
{cmd:groups()} option may be used to redisplay estimates that have already been
transformed by {cmd:devcon}. The {cmd:devcon} routine will work after most
estimation commands (see help
{help estcom}). Multiple equation models are
supported. Use the {cmd:equations()} option to specify the equation(s) to
be transformed. Note that {cmd:devcon} also transforms the variance-covariance
matrix of the estimates and that the usual post estimation commands such as
{help predict} or
{help test} may be used with the transformed estimates.
{p 4 4 2}The {cmd:devcon} command has two main benefits. First, it may be
very convenient to use {cmd:devcon} to quickly display the deviation
contrasts without having to change the coding of the variables and without
having to take further action to make the reference category's coefficient
visible. Second, the transformed estimates may be valuable for
use with some post-estimation procedures. In fact, {cmd:devcon} was
originally developed for use with with the Oaxaca-Blinder decomposition
(see help
{help oaxaca} if installed; the package
is available from the SSC Archive, type
{net "describe http://fmwww.bc.edu/RePEc/bocode/o/oaxaca":{bind:ssc describe oaxaca}}).
In this decomposition, the results for categorical
variables depend on the choice of the reference category (see, e.g., Oaxaca
and Ransom 1999). Applying the deviation contrast transformation to the
estimates before conducting the decomposition is one solution to this
problem (see Yun 2003).
{it:Technical note}
{p 4 4 2} The deviation contrast transform can also be applied to the
variables used to model an interaction between a categorical and a
continuous variable. The relevant continuous variable must be provided in
parentheses within the {cmd:groups()} option in such a case.
{title:Options}
{p 4 8 2} {cmd:groups(}{it:glist}{cmd:)} defines the dummy-variable groups.
If more than one group is specified, use commas to separate the groups.
Note that in each of the groups a variable reflecting the reference
category must be specified (i.e. the variable must exist in the data). If
the variables in a group represent interactions with a continuous variable,
specify the continuous variable in parentheses at the end of the group. The
usual shorthand conventions apply to the {it:varlist}s specified in
{it:glist} (see help
{help varlist}).
{p 4 8 2}{cmd:equations(}{it:numlist}{cmd:)} is relevant only for
multiple-equation models. It specifies the equation(s) to be transformed.
Use numbers to refer to the equations' positions in the model ({cmd:1} for
the first equation, {cmd:2} for the second, and so on). The usual shorthand
conventions apply to {it:numlist} (see help
{help numlist}). The default is {cmd:equations(1)}.
{p 4 8 2} {cmd:check}[{cmd:(}{it:#}{cmd:)}] checks the integrity of the
normalized estimates by verifying that the linear predictions from the
original estimates and the normalized estimates are equal for all
observations in the estimation sample. If the results do not pass the
check, an error message is issued and no results are returned. By default,
the check is performed using the models's first equation. To use another
equation, specify its number in parentheses. A failed check indicates that
the dummy variables used are not well defined (i.e. that the indicated
groups overlap or that at least one group has been omitted). In rare cases,
however, the results might fail the check even though the dummy variables
have been correctly defined ({cmd:devcon} uses the information in
{cmd:e(sample)} and, if available, {cmd:e(subpop)} to determine the sample
of relevant cases; situations may arise in which the sample would have
to be narrowed further).
{p 4 8 2} {cmd:nonoise} suppresses the display of the transformed estimates.
{p 4 8 2}{cmd:level(}{it:#}{cmd:)} specifies the confidence level, in percent
terms, for the confidence intervals of the coefficients; see help
{help level}.
{title:Example}
{p 4 4 2} Standard application with one categorical variable ...
{com}. sysuse auto
. generate rep1 = rep78 <= 3 if rep78 < .
. generate rep2 = rep78 == 4 if rep78 < .
. generate rep3 = rep78 == 5 if rep78 < .
. logit foreign mpg rep2 rep3, nolog
. devcon, groups(rep1 rep2 rep3){txt}
{p 4 4 2} ... and interactions with a continuous variable:
{com}. generate mpgrep1 = mpg * rep1
. generate mpgrep2 = mpg * rep2
. generate mpgrep3 = mpg * rep3
. logit foreign mpg rep2 rep3 mpgrep2 mpgrep3, nolog
. devcon, groups(rep1 rep2 rep3, mpgr* (mpg)){txt}
{p 4 4 2}Transforming OLS estimates for use with the Blinder-Oaxaca
decomposition ({cmd:oaxaca} is available from the SSC Archive):
{com}. reg lnwage educ expr expr2 single divorced if female==0
. devcon , groups(married single divorced)
. estimates store male
. reg lnwage educ expr expr2 single divorced if female==1
. devcon , groups(married single divorced)
. estimates store female
. oaxaca male female, detail
{txt}
{title:Methods and Formulas}
{p 4 4 2}Consider the model
y = a + b_1*D_1 + b_2*D_2 + e
{p 4 4 2}where "a" is the constant and "e" is the error. D_1 and D_2
are two 0/1 dummy variables representing a polytomous variable with three
categories. Alternatively, the above equation can be formulated as
y = a + b_1*D_1 + b_2*D_2 + b_3*D_3 + e
{p 4 4 2}with b_3 constrained to zero and D_3 being the indicator for the
(omitted) reference category. Now define c as
c = (b_1 + b_2)/3
{p 4 4 2}and let
a' = a + c
b_1' = b_1 - c
b_2' = b_2 - c
b_3' = b_3 - c = -c
{p 4 4 2}{cmd:devcon} then reports the equation
y = a' + b_1'*D_1 + b_2'*D_2 + b_3'*D_3 + e
{p 4 4 2}More generally,
c = (b_1 + b_2 + ... + b_{k-1}) / k
{p 4 4 2}for a k-category variable.
{p 4 4 2}The transformation can also be applied to interaction terms. Consider
the model
y = a + b_1*DX_1 + b_2*DX_2 + d*X + e
{p 4 4 2}where X is a continuous variable and DX_1 and DX_2 are the
interaction terms, i.e. DX_1 = D_1*X and DX_2 = D_2*X. The deviation
contrast transformation is then
y = a + b_1'*DX_1 + b_2'*DX_2 + b_3'*DX_3 + d'*X + e
{p 4 4 2} where
b_1' = b_1 - c
b_2' = b_2 - c
b_3' = b_3 - c = -c
d' = d + c
{p 4 4 2}{cmd:devcon} also transforms the variance-covariance matrix of the
coefficients, applying the general formula for the
variances and covariances of weighted sums of random variables
(see Mood et al. 1974:179).
{title:References}
{p 4 8 2}
Mood, A.M., F.A. Graybill, D.C. Boes (1974). Introduction to the Theory
of Statistics, 3. edn. New York: McGraw-Hill.{p_end}
{p 4 8 2}Oaxaca, R.L., Ransom, M.R. (1999). Identification in Detailed Wage Decompositions.
The Review of Economics and Statistics 81: 154-157.{p_end}
{p 4 8 2}Yun, M.-S. (2003). A Simple Solution to the Identification
Problem in Detailed Wage Decompositions. IZA Discussion Paper No. 836.{p_end}
{title:Author}
{p 4 4 2}Ben Jann, ETH Zurich, jann@soz.gess.ethz.ch
{title:Also see}
{p 4 13 2}
Online: help for
{help regress},
{help estimates},
{help oaxaca} (if installed)