I've just finished reading "Methological Errors in Medical Research" by Bjorn Andersen. It was recommended by Ben Goldacre in "Bad Science". I am on the lookout now for things that would help enable me critique and criticise research studies better and this sounded like it might be useful. Here's the official blurb: Table of contents: I found it an interesting read. He seems a very intelligent guy who can find subtle problems with papers that have been published. A lot of the examples don't require any statistical knowledge, while a smaller percentage would require some basic knowledge of statistical tests (but perhaps he explains what is involved enough for previous knowledge not to be necessary; I have some knowledge myself so hard for me to say). Apart from a handful of references (which could be skipped over), it doesn't require knowledge of any third level mathematics. A percentage of examples did cause me to look at a medical dictionary. (Some of that was probably not necessary: I like to broaden my knowledge of medical matters) I gradually read it over a few months. Although not as dense as a mathematical textbook, it is still somewhat dense. I found myself having to reread it bits of it. There were still some bits that I didn't understand. I might come back to it in 5-10 years for revision and to see if I understand the bits I didn't get the first time. Overall I'm pleased that I read it. Unfortunately I'm not sure if there are that many cheap copies around to buy but perhaps people can request it from libraries if they are interested.