0
$\begingroup$

I am looking for a book (containing a chapter)/pdf on Probit and Logit Analysis, with Logit Regression. I read Casella Berger but I find the topic is rather poorly written.

Also, some universities provide lecture handouts that are available online. Can you kindly give me a link to those? I have tried several such pdf's but the main problem is, none of them explains the theory. All those are presentation-like documents and it is hard to understand what they are trying to say.

Thank you.

$\endgroup$
3
$\begingroup$

I agree that the chapter of Casella and Berger is poorly written, not to mention quite short. Depending on your level, a good book on the topic is either "Introduction to Categorical Analysis" or its more advanced older brother Categorical Data Analysis, both written by Alan Agresti. I believe there are free pdf versions available if you look hard enough.

If you google his name, a bunch of relevant lecture notes will also come up. Give it a try.

$\endgroup$
  • $\begingroup$ Thanks. I am a first year undergrad in Stats. I know linear regression but have ABSOLUTELY no base in probit and logit analysis. However, in Linear regression, I do know a lot. $\endgroup$ – Landon Carter Jun 6 '15 at 14:23
  • $\begingroup$ @LandonCarter Go for the first one then. It starts from basic statistical inference, proceeds to contingency tables and afterwards examines in detail logistic regression with some references to probit as well. $\endgroup$ – JohnK Jun 6 '15 at 14:25
  • 1
    $\begingroup$ @LandonCarter I have studied Casella and Berger quite closely and it is very hard to follow. I wonder why it is considered the benchmark for intermediate mathematical statistics, so many better books out there. Feller is historic because of his discoveries, e.g. stars and bars, I agree nevertheless that he is hard to read. I have not read Rao's book because I suspect it will be the same. A good modern econometrics book like Hayashi is all one needs for the linear model. $\endgroup$ – JohnK Jun 6 '15 at 14:38
  • 1
    $\begingroup$ For Time Series I have only studied Hamilton and Enders in the past. Hamilton is considered a very solid reference but he can be quite advanced. Enders is more applied and presents several motivating examples from economics. For mathematical statistics, I wholeheartedly suggest "Introduction to Mathematical Statistics" by Hogg and Craig. It is less advanced than Schervish but if you are first year I guess you would appreciate the intuition behind the results, no? These two guys have also made important contributions to Statistics, e.g. in quadratic forms. $\endgroup$ – JohnK Jun 6 '15 at 14:49
  • 1
    $\begingroup$ @LandonCarter No worries. Take a look at Agresti and Hogg and Craig. Don't worry yet about TIme Series. $\endgroup$ – JohnK Jun 6 '15 at 14:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.