Gearing up for statistics PhD program

[danthechub]

Hi all,

I will be attending a top 10 stats PhD program in the fall, and am embarrassed to say that I have very limited experience in statistics. I got interested in stats because of my mathematical interests in probability, analysis, stochastic processes, and the few stats courses I have taken. I think the programs found my math background attractive and I might perform better than most on, say, qualifying exams in measure-theory based probability. However, I would like to get very comfortable with the basic ideology and logic behind statistical methods, as well as survey a lot of different ideas in the field to get a sense of what direction I might be interested in.

Does anyone in a stats PhD recommend any books or resources to help me make the transition to stats? Any other advice on preparing for entering the program would be appreciated too. Thanks much!

[cyberwulf]

Have you taken a course in mathematical statistics? If not, you might consider grabbing a copy of Casella & Berger and reading though it. It's pretty much the canonical book for a first-year graduate math stat course.

[statprospect]

You might also want to bone up on some basic statistical programming. You could try reading through The Art of R Programming by Norman Matloff. He has a copy of the pdf freely available on his website (http://heather.cs.ucdavis.edu/~matloff/132/NSPpart.pdf).

[app_stat]

+1 to the Casella and Berger recommendation. We used that as our primary book for our year-long Probability and Mathematical Statistics sequence, and I still refer to it regularly. However, if you plan to take a more theoretical approach in your program, you might want to consider Rick Durrett's books on probability and stochastic processes. The former begins with a chapter on measure theory, which should give you a sense of his approach to the rest of the topics. (Note that measure theory is not really addressed in Casella and Berger's treatment of probability.)

Two great books for self-study:

Linear Models in Statistics (Rencher & Schaalje): This book takes you through all the standard linear models (and then some) from a linear algebraic perspective (standard at the graduate level). Clear exposition and the inclusion of nearly all solutions distinguish this book from others on the same subject. If you have any interest in econometrics, this will be a good primer that should allow you to dive right into Hayashi's and Wooldridge's respective textbooks.

A Course in Large Sample Theory (Ferguson): This was recommended by a professor as a complementary text to Casella and Berger. There's a more thorough treatment of asymptotics and convergence concepts here, plus this book also includes solutions in the back.

Congrats on your acceptance!