Description: The problem was to determine how comparable is comparable improvement for campuses in Texas. An alternative strategy for determining comparable improvement was developed using 2000 comparable improvement data provided by the Texas Education Agency for 2,403 elementary campuses. Comparable improvement is a measure that shows how student performance has changed from one year to the next and then compares that growth to 40 schools that are demographically most similar to the target school. Instead of using the most dominant characteristic as in the current process, the percent of students in poverty was the initial sorting characteristic. The impact of sorting by poverty was reviewed in four areas: 1.) the impact on quartile placement, 2.) the TLI average growth for the comparison group, 3.) the award eligibility, and 4.) the changes in comparison group composition. No practical significant difference was found for research questions 1 and 3, however, a practical significant difference was found in group average TLI growth for math and in the comparison group composition. Overall, the alternative process had the greatest impact on campuses with 40-80% poverty. Three possible factors may have influenced the results. First, the middle poverty campuses had the most change in comparison group as found in question 4. Second, the interaction between the middle poverty campuses and the alternative process could have been fueled by the removal of the 1,295 campuses with poverty as the dominant characteristic in current system. Third, the high correlation between poverty and ethnicity may have limited any impact of the alternative process.
Date: December 2001
Creator: Holland, Vicki Gay