Blake Miller's column, "SATs not indicative of success in college." demonstrates two things: First, she has no understanding of what correlation is, how it is indexed, or interpreted, and second, that she is blithely unaware of a large literature, spanning decades, concerned with the association between test performance and achievement. Correlation is a measure of linear association between two variables in a sample. For SAT scores and college grades it is modest at best, often around .4. Certainly there are many factors that contribute to college success, which are not indexed by tests such as the SAT, and there is no dispute about that.

It is certainly legitimate to raise the issue of whether SAT scores should be used as one of the selection criteria for college admissions. But to state that there is no correlation between performance and SAT scores is simply nonsense.