10 August 2009

Measuring Educational Growth in Colorado

This year Colorado's annual reveal of CSAP results (state standardized testing) has focused on a move from looking at success or failure of individual students and the rate of success of individual students within that school, to a value added approach.

Under the old system, students were graded advanced, proficient, partially proficient, or unsatisfactory in each subject area, and schools received excellent, high, average, low, or unsatisfactory grades based upon the percentage of students in each category. Eleven years of testing with the old regime showed that poor kids did poorly and that schools with lots of poor kids did especially poorly.

The new regime focuses on the "Colorado Growth Model" which compares a child's improvement over a year to others with the same starting point on a percentile basis. High percentiles show more value added by the school. The analysis also looks at the likelihood that a student "will reach proficiency in reading, writing or math within three years, or by the time they are in 10th grade."

In reading, only 34 percent of deficient students will catch up in three years. That leaves 74,264 students not achieving proficiency out of an estimated 470,000 students in Colorado.

In writing, only 29 percent of students will catch up, leaving 116,115 students from grades four through 10 not writing at grade level.

In math, only 14 percent of students will catch up, leaving 134,503 deficient three years later.

Twenty-five schools, including four Denver high schools, were on pace to bring only 1 percent of their deficient students to proficiency in math in three years.

Schools with average improvement in the 65th percentile or better are rare.

Of the 1,958 Colorado schools in the growth model, 136 reach the 65th percentile or above in reading — or one in 15 schools.

In writing, there are 191 schools, or one in 10 schools.

And in math, there are 232 schools, or one in eight.

Only a handful of schools have reached those levels in each of the past three years, which would suggest the type of sustained improvement necessary to boost student achievement levels: 10 schools in reading, 25 schools in writing and 31 schools in math.

There are standouts, like the Denver School of Science and Technology, where "For the past three years, DSST students have averaged 75 percent median student growth in all subjects, meaning the typical DSST student demonstrated more growth than 75 percent of other similar-scoring students statewide."

The measure shows that many Denver schools, some of which look bad on conventional measures because their students are so poor and start out with weak academic ability, are among the best in the state in terms of improvement. For example,

Beach Court Elementary School — a 350-student school near Federal Boulevard and Interstate 70 in northwest Denver.

Half of Beach Court's students are fluent Spanish speakers, and 96 percent are eligible for federal meal benefits — an indication of poverty.

Yet the school had the state's highest median growth rate in writing, the 11th-highest in math and 12th-highest in reading.

Other Denver standouts in reading and/or writing were Bromwell, Asbury, Lincoln, Steck, and University Park elementary schools, and West Denver Prep Middle School and KIPP Sunshine Peak. Four Denver high schools showed notable math improvement: Manual, Lincoln, North and Montbello.

No comments: