In general I'm wondering if there's any danger in studying from older edition statistics books. The undergrad material seems to have remained unchanged throughout the last century, but I'm looking for 'the next step', which makes me wonder if older editions will be modern enough.

For example, Applied Multivariate Statistical Analysis, Johnson/Wichern seems to be a book very much in my taste and I wouldn't mind saving $60 on a used copy from 1992. Would this have any drawbacks when self studying?


Look in the preamble to later editions to see what's changed & whether you care very much. There may only be a couple of things, which you could look up elsewhere. To be honest it's not uncommon for some modern technique or other to be shoe-horned in when it's treated better elsewhere. If you decide on an older edition try to supplement it with review articles that describe the state of the art & give the more recent references that will be lacking.

  • $\begingroup$ An interesting related question is how many times older editions have been shown to contain demonstrably incorrect material. $\endgroup$ – rolando2 Apr 19 '14 at 13:27
  • $\begingroup$ @rolando2: Outright errors are more likely in earlier printings; so you might be safer with the 7th printing of a 2nd edition (or mature errata) than the 1st printing of a 3rd edition. The inclusion of now-deprecated methods (& omission of modern alternatives) is something I think you have to assess for the particular book in question. $\endgroup$ – Scortchi - Reinstate Monica Apr 22 '14 at 8:57

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.