There is certainly a glut of books out there about why people reject science, embrace superstition/pseudoscience, or both, and I've read a few of them over the past month or so (The Mismeasure of Man, Discarded Science, Science Talk, and Why People Believe Weird Things are the ones I read) to see if I could get a better handle on things like belief in witchcraft, UFOs, and Holocaust denial. Unfortunately, many of the books were disappointing, being good sources for basic information but either fizzling out or generally lacking in any sort of style or conclusion, and I'm sorry to say that Michael Shermer's Why People Believe Weird Things falls into this category as well. Perhaps I'm holding Shermer's book to an unreasonable standard, Carl Sagan's The Demon-Haunted World and the posthumously published Varieties of Scientific Experience having a much greater effect on me intellectually than most other works I've come across, but even if this is so I still think Why People Believe Weird Things shows some promise but never fully gets off the ground.
Rather than starting from the beginning, my biggest gripe about the book is about the end; it simply stops without warning or summation, the new included chapter seemingly being out of context from the rest of the work and shoved in the end pages. I don't know who to blame for this but it is certainly lazy and made for a very disappointing end to the book. As for the rest of the work, it's not a bad collection of claims of people who "believe weird things" with complimentary chapters dealing with the refutation of the claims, and for this reason the book can serve as a handy reference or resource should you encounter strange claims. The problem with this setup is that a significant amount of the text is arranged in lists, which to me is rather lazy writing that does not require the other to really connect the ideas being presented. I can understand the utility of such an approach, but it does take away some enjoyment when reading the book.
Outside of editing/style errors, there are a few points that jumped out at me as being a bit undeveloped. Shermer's idea of a "belief engine" in the human mind is briefly mentioned in the introduction and quickly passed by, although this is for the better as it seemed to dabble in evolutionary psychology a bit (which I don't want to discount as a whole but reminds me of hyper-adaptationist arguments where correlation is confused with causation). Likewise, Shermer makes a distinction between two types of errors early on in the book, the acceptance of a falsehood and the rejection of a truth, but what he doesn't do is tie the two together. The weird beliefs discussed in the book require both a rejection of truth and the acceptance of a false belief, one usually preceding the other. In the case of creationism, a truth (evolution) is rejected, creationism filling in to replace it. In other cases, perhaps in some instances of Holocaust denial, a fact is rejected but something must fill the void and so a falsehood is pieced together, showing that the two types of mistakes are linked together and seldom (if ever) really exist by themselves.
All of this is not to say that Shermer's book is a "bad" book or of no value; it provides a useful overview of untenable claims and promotes the value of skeptical inquiry (the author taking time to distinguish it from atheism and the idea that skeptics are just old cranks that are angry at the world). The personal touch that Shermer adds, either about his somewhat more gullible days or his encounters with "true believers" of various stripes, is effective when used as well, although the personal narratives usually simply serve as an introduction to a topic and are not fully carried through. In the end, the book was a fair overview of various beliefs that don't hold any water at all, but I would by lying if I said it held a candle to The Demon-Haunted World.
I read Shermer's book a while back, and now I don't remember it in detail, but recall coming away with the impression that while it did go into some depth about particular beliefs, it didn't really do that much toward answering the general question posed by its own title.
I'll agree with Moopheus on this one; a more accurate title might have been, Here Are Some Weird Things People Believe!
On a somewhat related note, I find it interesting that so many people have thrown Courtier's Replies at Richard Dawkins, whereas so few have taken Shermer to task for ignorance of history and archaeology, subjects infinitely more grounded in empirico-rationalism than theology. In Fighting Words (2005), Hector Avalos makes a good case that Shermer's Science of Good and Evil (2004) comes up shallow in the research department.
Of course, I was a little displeased with Fighting Words itself: at least in the printing I read, it really could have used a more watchful copy editor. While I would still recommend the book, and quite warmly, a general "glitchiness" (mostly at the punctuation level) made it seem like a step in the publishing process had been skipped.