Charles Murray on Regression Analysis

I have questioned Murray on his description (and reach) of regression analysis. Here is the quote from the Bell Curve:

 

The basic tool for multivariate analysis in the social sciences is known as regression analysis. The many forms of regression analysis have a common structure. There is a result to explain, the dependent variable. There are some things that might be the causes, the independent variable. Regression analysis tells us how much each cause actually affects the result, taking the role of all other hypothesized causes into account — an enormously useful thing for a statistical procedure to do, hence its widespread use. (The Bell Curve, p. 122)

The sentence of dispute is: “Regression analysis tells us how much each cause actually affects the result.”

Share this post:

Recent Posts

One Comment

  1. Joseph Lowe says:

    I won’t pretend to be a statistician. To date, I’ve only completed 8 statistics courses, 2 undergrad, 2 masters, 4 doctoral, in support of other studies. Suffice it to say, these courses taught me the basics and how to use and interpret results from software like SPSS. In other words, I know enough to get me into trouble.

    It’s my understanding that linear regression analysis only shows strength and direction of a relationship, not causation. Is it possible that the sentence in question above is a grammatical or editing error? Has it been addressed in previous debates or ignored, to your knowledge? Wondering aloud, why on earth would such a gregarious error such as this make it to print if it weren’t in error? I still consider myself an idiot when it comes to statistics and this immediately leaped out to me, only after you brought attention to it. If a basic premise such as this forms the foundation of Murray’s argument, how many other premises are based on faulty assumptions? This will be a very interesting analysis.
    Warmest regards,
    Joey Lowe

Leave a Comment