#StackBounty: #r #regression #logistic #chi-squared #log-likelihood chi-square goodness-of-fit and R-square measures from the fixed-eff…

Bounty: 50

I am trying to get the chi-square goodness-of-fit and R-square measures from the following fixed-effect logistic regression using ‘feglm’ function.
However, I find very limited information to even check this.

    > regress=feglm(Y ~ X1+X2+X3+X4+X5+X6*X10+X7+X8+X9+X11| Firm+Time, data=DATA, family=binomial(link="logit"))
    > summary(regress)
    binomial 

    Y ~ X1 + X2 + X3 + X4 + X5 + X6 * X10 + 
        X7 + X8 + X9 + X11 | Firm + Time

    l= [127, 15], n= 14139, deviance= 9891.112

    Structural parameter(s):

               Estimate  Std. error z value  Pr(> |z|)    
    X1        -7.006e-02  3.990e-03 -17.560   < 2e-16 ***
    X2        1.473e+00  1.047e-01  14.077   < 2e-16 ***
    X3        -9.105e-02  2.691e-02  -3.384  0.000715 ***
    X4        -2.896e-04  3.294e-05  -8.791   < 2e-16 ***
    X5        1.223e-01  4.557e-03  26.848   < 2e-16 ***
    X6        1.154e-01  2.267e-01   0.509  0.610699    
    X10       -6.273e-03  2.387e+00  -0.003  0.997903    
    X7        2.663e-02  1.192e-02   2.234  0.025453 *  
    X8        2.940e-01  9.002e-02   3.266  0.001092 ** 
    X9        4.115e+00  1.080e-01  38.103   < 2e-16 ***
    X11       1.115e-03  3.442e-01   0.003  0.997415    
    X6:X10    3.344e-02  2.533e-01   0.132  0.894962    
    ---
    Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
    ( 6244 observation(s) deleted due to missingness )

I suppose I will need at least loglikelihood, residual deviance, etc. values to even calculate the chi-square values and R-squares which I cannot find from the above result.

May I get help on this?


Get this bounty!!!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.