Monday, May 05, 2003


THE “PAIN” OF MIDDLE-AUSTRALIA

Craig McGregor has written a rather sycophantic review of Michael Pusey’s latest book The experience of middle Australia. The book’s basic thesis is that economic reform has caused great anguish to middle-class Australians. Both the book and the review have such a strong and evident Leftist bias that I was going to leave it to other conservative bloggers to “fisk” both as I like to put my efforts into making my own points rather than demolishing somebody else’s obvious nonsense. Since I know the author of the book quite well, however, China hand has persuaded me that I should put up a few notes.

The first laugh I had when I read the review was: It's a bold and formidable achievement (with endorsements from Noam Chomsky and others on the cover). That should be enough for most people who are wary of bias.

But anyway: Up to 1983 I had an office next to Michael Pusey at the University of N.S.W. and found him to be a most pleasant, bright and friendly fellow. He was however a keen Marxist at that time so his claim now that he is a "middle-of-the-road social democrat" is at least not the whole truth.

I have not read the book nor do I intend to as I gather that it is based on face-to-face interviews and anybody who knows anything about social research knows how much room for observer bias that data gathering of that sort enables. And that the researcher was biased McGregor clearly tells us. He says that Pusey is “an absolutely committed critic of economic and structural reform and globalisation, using his research results to mount a passionate condemnation of the social impact of these reforms.” Pusey gets a bit much even for for McGregor in the end -- he says that the epilogue to the book “descends into an anti-American rant. "The United States has gone mad," he writes”.

That researchers find what they expect to find is known among psychologists (Pusey is a sociologist) as The Rosenthal effect. It is an amazingly powerful effect and not easily avoided. One reason why psychometricians such as myself rely heavily on multiple choice questionnaires is that it enables the researcher to have NO contact with the people being studied -- thus precluding a major source of bias. In most of my research career, until the numbers came out of the computer I had no idea what I was going to find. Questionnaires have their limitations, too, of course, but that is another story.

*************************************

No comments: