Based on recent information provided by the Accountability Technical Advisory Committee’s (TAC) recommendations to ISBE, it is likely that ISBE will move away from linear regression toward Student Growth Percentiles (SGPs) as a means to measure student growth under the new school improvement and accountability system.
I’m getting a lot of questions related to linear regression and how the Illinois State Board of Education (ISBE) may implement linear regression to measure student growth under the new school improvement and accountability system. As a result, I decided to synthesize information as it becomes available.
The fundamental purpose of setting school improvement goals is to assess the effectiveness of improvement efforts. The utility of the school improvement process rests on the inferences one can draw from meeting a goal. For example, a typical school improvement goal may be “increase the percentage of students meeting standards in grade 4 reading by 5%.” While this type of goal is specific and concise, the results will provide little to no information as to the effectiveness of improvement efforts. Why? Because increasing grade 4 reading by 5% is likely an arbitrary goal that is unrelated to school performance and not evidence-based.