The Nature and Origins of Mass Opinion (Cambridge Studies in Public Opinion and Political Psychology)

If you need a primer on public opinion research, this isn't it. But if you want to read cutting-edge work, and you have a good grasp of statistics, dive in.

But this does not prove that people are uninformed, but that standard polling is woefully inadequate in measuring all of the idiosyncrasies of the human mind, and one's opinions on complicated political matters. This book gets off to a great start by illuminating such fallacies. The first few chapters are strong enough to earn this book accolades as a poli-sci landmark. But after proving that public opinion is perhaps unmeasurable (at least accurately), Zaller then spends the rest of the book measuring it himself anyway, with self-designed statistical models. However, it becomes difficult to tell whether he is using the results of his measurements to gain insight into actual public opinion, or merely to prove the viability of his own statistics.

By the middle of the book Zaller has embarked on a tedious and uninspiring academic exercise in statistical modeling that adds little to the points that were already made convincingly early on. Unnecessarily complex statistical equations are piled on mundanely, along with unenlightening charts and graphs. Like many other political science writers, Zaller has focused on pleasing his colleagues who are more likely to be impressed by repetitive evidence and windy statistics, while forgetting about the informed general reader who may just find the big-picture conclusions fascinating. That makes the majority of this book somewhat disappointing, rather like standing on a mountaintop on a cloudy day. You know the view (i.e. insights into the political knowledge of the masses) could be tremendous, if it wasn't obscured by clouds (i.e. academic tedium). True knowledge can be gained by disregarding the dusty science of this book and concentrating on the greater insights that lie beneath. [~doomsdayer520~]