R help - May 2012 - Hypothesis Testing using Wald Criterion for two regression models with dummy variables

I have two models, controlled by dummy variables to see if the models can be
combined into one model with similar intercepts and slopes. Has anyone tried
to conduct this type of test in R. I am utilizing the econometric idea of
hypothesis testing through the hypothesis of coincidence. I have tried to
run an anova with test of Chisq, but I am not sure what the results are
telling. In addition, I used the rms package with a lrm model in an anova
test, again I am not sure what the results are telling me:

Try 1
anova(H,Ha,test="Chi")
Analysis of Variance Table

Model 1: logload ~ logflow
Model 2: logload ~ dummy + logflow
  Res.Df    RSS Df Sum of Sq  Pr(>Chi)    
1     17 4.6742                           
2     16 2.6314  1    2.0428 0.0004245 ***
---
Signif. codes:  0 ?***? 0.001 ?**? 0.01 ?*? 0.05 ?.? 0.1 ? ? 1 

Try 2:
fit<-lrm(logload~logflow+dummy)
> anova(fit)
                Wald Statistics          Response: logload 

 Factor     Chi-Square d.f. P     
 logflow    17.56      1    <.0001
 dummy       5.22      1    0.0224
 TOTAL      18.03      2    0.0001

Can anyone help me with this?
Thanks!

--
View this message in context:
http://r.789695.n4.nabble.com/Hypothesis-Testing-using-Wald-Criterion-for-two-regression-models-with-dummy-variables-tp4601582.html
Sent from the R help mailing list archive at Nabble.com.

On May 1, 2012, at 22:01 , meredith wrote:
> I have two models, controlled by dummy variables to see if the models can
be
> combined into one model with similar intercepts and slopes. Has anyone
tried
> to conduct this type of test in R. I am utilizing the econometric idea of
> hypothesis testing through the hypothesis of coincidence. I have tried to
> run an anova with test of Chisq, but I am not sure what the results are
> telling. In addition, I used the rms package with a lrm model in an anova
> test, again I am not sure what the results are telling me:
> ....
> Can anyone help me with this?
Not unless you say what you are trying to do... As far as I can tell, you are
(A) using a chi-square test for a standard linear model (Gaussian response) and
(B) fitting a logistic regression model to the same response, which assumes that
it is a proportion or binary response. Neither makes any sense to me.

-- 
Peter Dalgaard, Professor
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

Meredith:

You are clearly out of your depth. Get local help. R-help is, err....,
an R help list, not a resource for remote statistical consulting.
Although, I admit, there is often some overlap.

-- Bert

On Wed, May 2, 2012 at 5:58 AM, peter dalgaard <pdalgd at gmail.com>
wrote:
>
> On May 1, 2012, at 22:01 , meredith wrote:
>
>> I have two models, controlled by dummy variables to see if the models
can be
>> combined into one model with similar intercepts and slopes. Has anyone
tried
>> to conduct this type of test in R. I am utilizing the econometric idea
of
>> hypothesis testing through the hypothesis of coincidence. I have tried
to
>> run an anova with test of Chisq, but I am not sure what the results are
>> telling. In addition, I used the rms package with a lrm model in an
anova
>> test, again I am not sure what the results are telling me:
>> ....
>> Can anyone help me with this?
>
> Not unless you say what you are trying to do... As far as I can tell, you
are (A) using a chi-square test for a standard linear model (Gaussian response)
and (B) fitting a logistic regression model to the same response, which assumes
that it is a proportion or binary response. Neither makes any sense to me.
>
> --
> Peter Dalgaard, Professor
> Center for Statistics, Copenhagen Business School
> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
> Phone: (+45)38153501
> Email: pd.mes at cbs.dk ?Priv: PDalgd at gmail.com
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.


-- 

Bert Gunter
Genentech Nonclinical Biostatistics

Internal Contact Info:
Phone: 467-7374
Website:
http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm