[R] Likelihood ratio test using R

C.H. chainsawtiney at gmail.com
Wed Nov 5 08:45:11 CET 2008


For the third one:

?anova.glm

test=Chisq will be LRT.

For the first two, you can have the answer from ordinary stat book.


On Wed, Nov 5, 2008 at 1:11 PM, Maithili Shiva <maithili_shiva at yahoo.com> wrote:
> Hi!
>
> I am working on the Logistic Regression using R. My R script is as follows
>
>
> ONS <- read.csv("ONS.csv",header = TRUE)
>
> ONS.glm <- glm(Y ~ Age1+Age2+Sex+Education+Profession+TimeInJob+ TimeInCurrentJob+OwnResidence+Dependents+ApplIncome+FamilyInco+IIR+FOIR+YearsAtBank+SavingsAccount+CurrentAccount, family = binomial(link = logit), data=ONS)
>
> summary(ONS.glm)
>
>
> OUTPUT (RESULT) is as follows -
>
> Deviance Residuals:
>    Min       1Q   Median       3Q      Max
> -1.7888  -0.6910  -0.3220  -0.2179   3.1608
>
> Coefficients:
>                   Estimate Std. Error z value Pr(>|z|)
> (Intercept)      -1.454e+00  1.380e-01 -10.532  < 2e-16 ***
> Age1              2.313e-01  9.902e-02   2.336   0.0195 *
> Age2             -6.648e-01  5.209e-02 -12.761  < 2e-16 ***
> Sex              -7.012e-01  7.325e-02  -9.572  < 2e-16 ***
> Education        -5.271e-01  7.963e-02  -6.619 3.61e-11 ***
> Profession       -3.512e-01  4.908e-02  -7.155 8.38e-13 ***
> TimeInJob        -3.107e-01  1.980e-01  -1.569   0.1166
> TimeInCurrentJob  2.752e+00  1.895e-01  14.521  < 2e-16 ***
> OwnResidence      1.608e+00  1.858e-01   8.654  < 2e-16 ***
> Dependents       -4.066e-01  1.867e-01  -2.178   0.0294 *
> ApplIncome       -6.401e-02  4.319e-03 -14.820  < 2e-16 ***
> FamilyInco        7.148e-02  7.389e-03   9.674  < 2e-16 ***
> IIR               5.765e-02  1.353e-02   4.262 2.03e-05 ***
> FOIR              7.383e-07  1.056e-07   6.990 2.74e-12 ***
> YearsAtBank      -1.926e-07  3.519e-08  -5.473 4.42e-08 ***
> SavingsAccount   -2.083e-07  1.785e-07  -1.167   0.2432
> CurrentAccount   -1.550e+00  1.107e-01 -14.008  < 2e-16 ***
> ---
> Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
>
> (Dispersion parameter for binomial family taken to be 1)
>
>    Null deviance: 39300  on 42499  degrees of freedom
> Residual deviance: 32534  on 42483  degrees of freedom
> AIC: 32568
>
> ____________________________________________________________
>
> Question No - (1)
>
> What does this means?
>
> Deviance Residuals:
>    Min       1Q   Median       3Q      Max
> -1.7888  -0.6910  -0.3220  -0.2179   3.1608
>
>
> Question No - (2)
>
> What does this means?
>
> Null deviance: 39300  on 42499  degrees of freedom
> Residual deviance: 32534  on 42483  degrees of freedom
>
>
> And this is the last question
>
> Question No - (3)
>
> To test Ho : All regression coefficients are ZERO, I need to use Likelihood Ratio Test (G) and the related p value. When I use minitab, I get value of G and related p, but using R language, how do I get value of G and p value?
>
> _____________________________________________________________
>
> I am sincerely sorry for raising these many questions? But I am sure there will be many like me who will also be immensely benefited by the replies, I may get pertaining to these questions.
>
>
> Thanking you all in advance,
>
> With warm regards
>
> Maithili Shiva
>
>
>
>
>
> ______________________________________________
> 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.
>



-- 
CH Chan
Research Assistant - KWH
http://www.macgrass.com



More information about the R-help mailing list