I am a soon-to-be physician. During my studies I have taken a class in biostatistics. I own Martin Bland's "An introduction to medical statistics", which was a required textbooks at the time, and Harvey Motulsky's "Intuitive biostatistics", which I've purchased on my own initiative based on positive reviews on Amazon. They served me well, but I'm interested in clinical research and I feel that I need to stop scratching the surface and delve deeper to understand statistics.

I'm looking for one or more textbooks I can use to self study. Ideally the textbook(s) should go beyond the basics, and explain modern statistical techniques used in biomedical research (e.g reading these forums I've seen a lot of mentions of Bayesian statistics, which seems to be the Next Big Thing). If possible, it should feature examples, and not assume a particular statistical package (unless it is R, that is the software I use).

I should probably mentions that my math skills are ... rusty. So feel free to list the kind of math skills required to understand the book(s) you are suggesting.

  • 1
    $\begingroup$ It's hard to make a suggestion when you haven't really said what topics you're already familiar with, but if you're familiar with multiple regression then I'd suggest learning about mixed-effects models, from something like amazon.com/Applied-Longitudinal-Analysis-Garrett-Fitzmaurice/dp/… $\endgroup$
    – mark999
    Aug 31, 2014 at 11:50
  • 6
    $\begingroup$ I would not recommend studying mixed effects models until fixed effects models are mastered and random effects are needed. $\endgroup$ Aug 31, 2014 at 13:00
  • 5
    $\begingroup$ Start with studying/refreshing algebra: biostat.mc.vanderbilt.edu/PrereqAlgebra $\endgroup$ Aug 31, 2014 at 13:02
  • 3
    $\begingroup$ Harvey's made some further suggestions for you so if you liked his book, his suggestions may help. $\endgroup$
    – Glen_b
    Sep 1, 2014 at 2:20

3 Answers 3


I'd strongly suggest you get Frank Harrell's Regression Modeling Strategies.

Possibly not as your only book - you'll might want something more elementary as well (though since you've already read some books, this may be covered) - but this book is full of practical and important information, and if you use R (which is also an important tool, I think), then it's doubly valuable --- but highly valuable whether you use it or not.

I am not in the medical area, but nevertheless I tell all my prospective research students (if they haven't already read it) to borrow it (and consider buying it). I insist that at the very least they read chapter 4, which I think is unmissable information for anyone using regression - since all my subsequent discussion with them will assume they know what it says there.


Here are three recommendations of books that focus on plain language explanations and practical suggestions for conventional (not Bayesian or computer intensive) methods but also include the mathematical details:

enter image description here Maxwell and Delaney

enter image description here Machin

enter image description here Glantz and Slinker

  • Vittinghoff. Regression Models in Biostatistics (There is a Stata emphasis that might be annoying. In addition the Second edition isn't as well written as the first.)
  • Steyerberg. Prediction Modelling
  • ? Frank Harrell. Regression Modeling Strategies. I've never read so cannot vouch for. Seems like a natural choice if you're using R. Bit pricey.
  • do research. less statistics involved than it seems.
  • 2
    $\begingroup$ charles - on Frank Harrell's book -- I strongly suggest you get hold of a copy and read it. If you only have time to read some of it, read chapter 4. $\endgroup$
    – Glen_b
    Sep 1, 2014 at 2:18
  • $\begingroup$ Thanks. I've been meaning to get to it. I'll have to find the time. I liked the Steyerberg book a lot. Narrower focus. But still seemed to cover the basics surprisingly well. Often better than Vittinghoff book. $\endgroup$
    – charles
    Sep 1, 2014 at 2:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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