Volume 3 Issue 1

Measuring How Well Jeffco Schools Improve Student Achievement

by Tom Coyne

There are a number of different tests which measure student achievement. Two main criteria can help you determine how well schools improve student achievement. The first is absolute achievement, as measured by the percent of students who score in the proficient or advanced category compared against state grade level standards. Students take the tests in grades 3 through 10 in reading, writing, and math; they are tested in science and social studies only once in elementary, middle, and high school. For example, the state’s John Irwin Awards are based on the percent of students at a school that are proficient or advanced in reading, writing, math, and science.

One issue with absolute proficiency is that students may have arrived at the school significantly behind, so just measuring absolute proficiency might not reflect the work this year’s team has accomplished. For that reason, Senate Bill 10-191, Colorado’s framework for evaluating teachers and principals, does not use absolute achievement as a criteria, because it would unfairly benefit those educators who are based at schools were children are already at grade level. It also runs the risk of committing the error that was so colorfully described by Barry Switzer, Oklahoma’s legendary football coach: “Some people are born on third base and go through life thinking they hit a triple.”

Instead, SB 191 uses a different criteria that aims to capture the value added by a teacher and school. This criteria is called “Median Growth Percentile.” A student’s growth percentile is a relative, rather than an absolute, metric. Essentially, it is calculated by subtracting his or her previous year’s TCAP scale score from this year’s score, and dividing the result by the starting score to standardize the change from one year to the next. These standardized scores are then further classified into percentiles that range from 0 to 99. The Median Growth Percentile is simply the midpoint growth percentile in any grouping of students.

Because it is a relative measure, MGP makes it easier to compare value added across schools. For example, the Median Growth Percentile for math at Evergreen and Jefferson high schools could in theory be exactly the same, even though the students had very different starting and ending TCAP scale scores. In this example, both schools could be said to have added the same amount of value to their students’ math achievement over the previous year.

I like to use a running race analogy to explain the difference between absolute and relative achievement metrics. Suppose your child comes home and tells you that she or he finished in the 90th percentile in a running race at school. This tells you that, compared to the other students who were on the same starting line as your child, he or she was relatively fast. But does that tell you that you should start thinking about college track scholarships? No, it does not. For that, you need to know your child’s absolute time in the race. 

Both metrics are important. But because a school cannot control the previous educational experience of the students who walk in the door, it makes more sense to use a relative metric like Median Growth Percentile to measure the effectiveness of schools.

On the K12 Accountability web site (www.k12accountability.org),  I have listed the top 10 highest performing schools, as measured by their three-year average MGPs in math, reading, and writing for students in the following categories: 

1. Non Low-Income General Ed: Students who are not eligible for free and reduced meals, and do not have either an Advanced Learning Plan or Individualized Education Plan (ALP/Gifted Education or IEP/Special Education) 

2. Low-Income General Ed: Students who are eligible for free and reduced meals, and do not have either an ALP or IEP

3. Non-Native Speakers: Students who are English Language Learners 

4. Non Low-Income Gifted: Students who are not eligible for free and reduced meals, and have and ALP but not an IEP.

Other groups of students had too few students at a school level to allow for meaningful analysis. Charter, Option and District run schools are all included. 

Listed in the chart are the math and reading results for Students Not Eligible for Free and Reduced Meals, Who Have Neither an ALP nor IEP.

Jeffco’s choice enrollment program offers parents a wide range of options for finding a good fit between what their child needs and what various schools have to offer. Nonetheless, the district has always been stingy when it comes to providing parents with the information they need to easily compare schools and make good decisions. To be sure, there are state-level resources like Colorado School Grades (http://www.coloradoschoolgrades.com), and the Colorado Department of Education’s School View website (http://www.cde.state.co.us/schoolview). 

However, as I have noted above, their methodologies have some important limitations, particularly if a parent is trying to identify those schools that add the most value to student achievement, and not just schools that are located in the most affluent ZIP codes. Hopefully, this analysis will help Jeffco parents make better school choice decisions for their children.

Another point that is sure to strike a lot of people, particularly after they have reviewed the data for all schools on the downloadable Excel spreadsheet at k12accountability.org, is that Jeffco has both very strong and very weak schools (including some whose reputations substantially exceed their results).

One of my great frustrations with the way Jeffco was run for over a decade under former Superintendent Cindy Stevenson was that there seemed to be a great reluctance to publicly compare schools’ performance. The district also lacked strong management processes for learning lessons from the best performers and systematically transferring them to other schools. 

Coyne chart.jpg

Just ask any of the teachers and principals at the schools listed here with MGPs of 60 or higher how often they received visits from the head office or from principals and teachers at other schools seeking to learn how they achieved their impressive results. And if your school has MGPs below 60, ask your principal and teachers how often they reached out to these high performing schools to solicit improvement ideas and coaching. Sadly, you won’t hear many positive responses.

And that’s the point. If we want to improve student achievement in Jeffco, this practice needs to change, even if increased use of school performance comparisons ruffles some feathers and produces more conflict. The inescapable truth is that substantial performance improvement requires substantial change, which inevitably produces conflict. If Jeffco’s primary goal is to avoid that conflict, then our student achievement performance will never improve, and we’ll continue to pay a billion dollars a year for mediocre results.

The entire data set, as well as results and rankings for all Jeffco schools can be downloaded as an Excel (.xlsx) file from K12accountability.org.

Tom Coyne is a political Independent. He chairs the Wheat Ridge High School Accountability Committee, is a member of Jeffco’s District Accountability Committee, and has worked on corporate performance improvement issues for more than 30 years.