Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Rép. : st: RE: Command -contrast- after -anova


From   "Martin Weiss" <martin.weiss1@gmx.de>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: Rép. : st: RE: Command -contrast- after -anova
Date   Tue, 9 Dec 2008 16:21:58 +0100

Line for the server...

I am not an expert on the internals of Phil`s program, but it was released
in 2008, so at least this takes you closer to Stata 10. David suggested this
package as well, I think your messages just crossed... 


HTH
Martin


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Herve STOLOWY
Sent: Tuesday, December 09, 2008 4:18 PM
To: statalist@hsphsun2.harvard.edu
Subject: Rép. : st: RE: Command -contrast- after -anova

Dear Martin:

I use Stata 10. I use the official Stata's commands + -contrast-
(downloadable from ssc).

Here is my output:

contrast return info risk, coeffs(1 1),  if round>7 & round<26 & offer==1 &
win==1

  +-------------------------------------------------+
  | info   __n   __mean   __res_sd   __coeffs   __a |
  |-------------------------------------------------|
  | Full     0        .          .          1     0 |
  +-------------------------------------------------+

Variable |     Obs     Contrast    Std. Err.       [95% Conf. Interval]
         |                  (L)       SE(L)
---------+-------------------------------------------------------------
  return |    2309            .           .               .           .

Hypothesis test of contrast = 0 (L = 0):  t =       ., P =       .

Hypothesis test of equal info means:      F =  10.973, P =  0.0009

Residual SD (with 2306 df) from analysis of variance = .02661463

There is no real error message but I get almost only dots.

I found another command -anovacontrast- (donwnloadable from
http://www.ats.ucla.edu/stat/stata/ado/analysis).

Should I use it preferably?

Best regards

Herve


***********************************************************
Professeur/Professor
President of the French Accounting Association (AFC)
HEC Paris
Departement Comptabilite Controle de gestion / Dept of Accounting and
Management Control
1, rue de la Liberation
78351 - Jouy-en-Josas
France
Tel: +33 1 39 67 94 42 - Fax: +33 1 39 67 70 86
mail: stolowy at hec dot fr
web: http://www.hec.fr/stolowy
>>> "Martin Weiss" <martin.weiss1@gmx.de> 09/12/08 15:20 >>>
Line for the server...

What is the error code that you get? 

- net describe contrast, from(http://fmwww.bc.edu/RePEc/bocode/c)- gives
1998 as the distribution date, and the command requests -version 4.0- when
run. The examples given are indeed defunct, but it seems that no one has
worked on them for at least a decade. 

Anyway, you should say if you use anything other than official Stata`s
commands in your posts...

HTH
Martin


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Herve STOLOWY
Sent: Tuesday, December 09, 2008 3:06 PM
To: statalist@hsphsun2.harvard.edu
Subject: st: Command -contrast- after -anova

Dear Statalisters:

I have some difficulties in using the -constrast- command (downloadable from
ssc) after -anova-.

Here is my -anova- command line with the output.

anova return info risk info*risk if round>7 & round<26 & offer==1 & win==1,
detail

  Factor        Value          Value          Value          Value
  ------------------------------------------------------------------------
  info          1 0            2 1           

  risk          1 0            2 1           

                           Number of obs =    2309     R-squared     =
0.0401
                           Root MSE      = .026537     Adj R-squared =
0.0389

                  Source |  Partial SS    df       MS           F     Prob >
F
 
-----------+----------------------------------------------------
                   Model |  .067894615     3  .022631538      32.14
0.0000
                         |
                    info |  .007341846     1  .007341846      10.43
0.0013
                    risk |  .045207121     1  .045207121      64.20
0.0000
               info*risk |  .010246209     1  .010246209      14.55
0.0001
                         |
                Residual |  1.62318282  2305  .000704201   
 
-----------+----------------------------------------------------
                   Total |  1.69107744  2308  .000732703   


In the -contrast- command, I use the following syntax:

contrast return info risk, coeffs(1 1),  if round>7 & round<26 & offer==1 &
win==1 

and it does not work.

I am almost sure that the figures after coeffs are wrong. To be honest, I do
not understand very well the help file of -contrast-. (I tried the examples
given in this file, based on auto.dta, but it returns an error message).

Does someone have some practice with -contrast-?

(I read the Stata manual and the differents threads in the Statalist. It
seems that the -contrast- command should avoid using the command -test- with
matrices after anova).

Best regards

Herve








***********************************************************
Professeur/Professor
President of the French Accounting Association (AFC)
HEC Paris
Departement Comptabilite Controle de gestion / Dept of Accounting and
Management Control
1, rue de la Liberation
78351 - Jouy-en-Josas
France
Tel: +33 1 39 67 94 42 - Fax: +33 1 39 67 70 86
mail: stolowy at hec dot fr
web: http://www.hec.fr/stolowy

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index