.-
help for ^xtmod^            (Version 0.9 20 February 2009)
.-

Analyze and display interactions based on time-series data
----------------------------------------------------------

        ^xtmod^ @varlist@ [if] [in] [, Controlvariables(@varlist@) Regopts(stri
> ng) 
        ARROWOFFset(#.#) ARROWXLENgth(#.#) ARROWYLENgth(#.#) PRECision(#.#) 
        GLS NOCONStant MAX Tests]

Description
-----------

        ^xtmod^ helps running multiple moderated regressions. It generates the 
        necessary interaction variables, calculates the coefficients, optionall
> y 
        displays tests, and displays the interaction. It is mainly a convenienc
> e 
        script that instrumentalizes individually available commands such as xt
> gls. 

        Warning: The program has not been tested with categorical variables. Pl
> ease 
        notify the author of bugs.


Options
-------

        ^controlvariables^(@varlist@) specifies control variables to be include
> d in 
        the regression
        
        ^regopts(string)^ allows the user to pass options to the regression com
> mand, 
        e.g., panels(heteroskedastic) in the case of xtgls

        ^arrowoffset(#.#)^ specifies how much the arrow head is offset from the
>  
        left-most point of the regression line. Ranges from 0.0 to 1.0 

        ^arrowxlength(#.#)^ specifies how much the equation text is offset to t
> he 
        right from the arrow's head. Ranges from 0.0 to 1.0 

        ^arrowxlength(#.#)^ specifies how much the equation text is vertically 
> offset 
        from the arrow's head. Ranges from 0.0 to 1.0 
        
        ^precision(#.#)^ specifies the precision used to display numbers in the
>  graphic

        ^gls^ tells ^xtmod^ to run ^xtgls^ instead of ^xtreg^

        ^noconstant^ tells ^xtmod^ that the constant equals zero. Can only be u
> sed 
        when gls has been specified
        
        ^max^ tells ^xtmod^ to plot the graphic using the minimum and maximum v
> alues 
        of the moderator as opposed to mean +/- standard deviation values (the 
> default)

        ^tests^ instructs ^xtmod^ to run post-hoc tests on simple slope differe
> nces 
        (compare Aiken and West, 1991)


Examples
--------

        . ^xtmod roa apcogs salesdiff rf, d(leverage industry?) r(panels(hetero
> )) arrowylen(0.9) gls nocons tests^
        . ^xtmod lndroa C_lnapcogs C_sales, tests^

References
----------

        Aiken, L.S. and S.G. West. "Multiple Regression: Testing and Interpreti
> ng Interactions. 
        1991. Thousand Oaks: Sage Publications.

        Stata FAQ, How can I understand a continuous by continuous interaction?
>  UCLA: 
        Academic Technology Services, Statistical Consulting Group. 
        from http://www.ats.ucla.edu/stat/stata/faq/concon.htm (accessed Februa
> ry 20, 2009).

        Stata FAQ, Part 1: How can I understand a 3-way continuous interaction?
>  UCLA: 
        Academic Technology Services, Statistical Consulting Group. 
        from http://www.ats.ucla.edu/stat/stata/faq/con3way.htm (accessed Febru
> ary 20, 2009).

Author
------

        Daniel Seifert, Ecole Polytechnique Fédérale de Lausanne (EPFL), Switze
> rland
        daniel.seifert@@epfl.ch

Also see
--------

 Manual:  ^[R] xtreg^, ^[R] xtgls^, ^[G] schemes intro^
On-line:  help for @xtreg@, @xtgls@, @schemes@