Ultimately, most methods for evaluating schools involve comparisons. Sometimes this involves comparing a school's performance on a statewide assessment, like the MEAP, against a standard of performane, such as Adequate Yearly Progress. Other times, we compare how well one school does against another, or a group of schools (PSAs) versus a different population (traditional public). These techniques range in their usefulness for school leaders and board members partly because many variables affect a school's academic performance.

But what if we could devise a method, or more accurately a model, to provide more accurate comparisons - those that are sensitive to the population of students the school serves? Such is the goal of regression analysis, a mathematical technique used to establish the relationship between two highly correlated variables.

Too much technical jargon already? Let's begin with an example away from education.