[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: Re: st: ask help for -fitstat- after regresion
Richard Williams,thank you！
you make it quite clear,and i can catch what i want.actually,i know i can find answers from Long&Freese's book,but i could not have access to it.it is quite a pity!
and anyone can tell where i can find some information about -fitstat-,if i want to learn more.any papers or other materials relative are appreciated!
======= 2004-03-21 15:17:25 您在来信中写道：=======
>At 01:00 PM 3/21/2004 +0800, rghuang wrote:
>>my question is as follows:
>Long & Freese go over this in their book if you happen to have access to
>it. Very quickly,
>>1.what dose D(66) means?and what it tells us?
>D = Deviance = -2 * Log-Lik Full Model. The 66 is the DF - # of parameters
>(Looks like you have 69 cases and 2 IVs.)
>>2.as we knows,aic and bic tell us how to choose the best model.the model
>>with small aic and bic is the better one.but what is the use of aic*n?and
>>what is bic'?how to caculate it?and what does it tell us ?
>AIC*n is, literally, AIC * N. In your case, 18.575 * 69 = 1281.675 (the
>slight difference is rounding error). Long and Freese say some people
>report this instead of AIC.
>BIC' is an alternative version of BIC. The more negative, the better.
>>Measures of Fit for regress of price
>>Log-Lik Intercept Only: -647.799 Log-Lik Full Model: -637.829
>>D(66): 1275.659 LR(2): 19.939
>> Prob > LR: 0.000
>>R2: 0.251 Adjusted R2: 0.228
>>AIC: 18.575 AIC*n: 1281.659
>>BIC: 996.208 BIC': -11.470
>>* For searches and help try:
>Richard Williams, Notre Dame Dept of Sociology
>OFFICE: (574)631-6668, (574)631-6463
>WWW (personal): http://www.nd.edu/~rwilliam
>WWW (department): http://www.nd.edu/~soc
>* For searches and help try:
= = = = = = = = = = = = = = = = = = = =
* For searches and help try: