.-
help for ^fitint^ (^Neville Verlander and Andre Charlett^)
.-
Fitting a regression model and testing each term
------------------------------------------------
^fitint^ regression_cmd yvar xvarlist [weight] [^if^ exp] [^in^ range]
[^,^ ^fac^tor^(^varlist^)^ ^twoway(^varlist [,varlist] ^)^
^noshow^ regression_cmd_options ]
where regression_cmd is one of
@clogit@ @nbreg@ @scobit@
@cloglog@ @ologit@ @stcox@
@cnreg@ @oprobit@ @streg@
@glm@ @poisson@ @tobit@
@logistic@ @probit@
@logit@ @regress@
Note: The yvar is not required for the stcox and streg commands but the data
need to be @stset@
Warning: In order to generate any interactions between continuous variables
this program uses the naming convention __x_y (double underscore) where x and y
denote the position of the two variables in the x variable list. In order for
this to work any pre-existing variables prefixed __ must be deleted or renamed.
The program will not run until this is done.
Description
-----------
Fitint fits a regression model and tests any twoway interactions and any
main effects not included in interaction terms. The appropriate test (LR of F)
is chosen depending upon the regression model being estimated.
Options
-------
^noshow^ suppress the regression table output
^fac^tor(varlist) define those xvarlist terms that are factors
^twoway^(varlist[,varlist]) define those xvarlist terms for which twoway interactions
are required. If more than two x-variables are listed then all possible twoway
interactions are generated. When a comma is used to separate x-variables lists then
all possible twoway interactions within each list are generated. e.g.
twoway(A B C, D E F) will produce the six interactions A*B A*C B*C D*E D*F E*F
regression_cmd_options are options appropriate to the regression command in
use. The ^cluster()^, ^robust^ and ^noconstant^ options are not available with
this command, if these options are issued the program will detected this
and exit.
Examples
--------
Suppose ^age^ is a continuous variable, ^sex^, ^auber^ and ^okra^ ^categorical^:
. fitint glm case age sex okra auber, show factor(sex auber okra)
twoway(auber age sex, auber okra) f(bin)
If the user wishes to use the ^logistic^ command, an example might be:
. fitint logistic case age sex okra auber, factor(sex auber okra)
twoway(auber age sex, auber okra) asis
Also see
--------
Manual: ^[U] 23 Estimation and post-estimation commands^,
^[U] 29 Overview of model estimation in Stata^
On-line: help for @lrtest@, @test@, @anova@; @clogit@, @cloglog@, @cnreg@,
@glm@, @logistic@, @logit@, @nbreg@, @ologit@, @oprobit@, @poisson@,
@probit@, @regress@, @scobit@, @stcox@, @streg@, @tobit@, @xi@, @lrdrop1@.