Colorado schools and the public are getting a more detailed picture of how students performed on the state’s new and more difficult annual assessments with the public release Friday of district- and school-level results.
However, state and district officials say the latest results from the PARCC tests fail to provide a complete picture of student achievement in Colorado because of low participation rates, especially among high school students. The high numbers of opt-outs, because they are concentrated at high-performing schools, all but certainly depressed the state’s scores.
Friday’s new data follows the much-anticipated release of statewide data last month. Those results found — as most people expected — that most Colorado students did not meet state expectations in English language arts and math.
And at first blush, district and school results follow familiar patterns, with middle-class suburban districts performing above the state’s average and school districts that serve mostly low-income students performing below the average.
State and district officials are emphasizing that the first PARCC scores serve as a baseline, a foundation on which students and educators can build as they continue to adapt to the Common Core standards in language arts and math.
In language arts, 40 percent of Colorado students rated as meeting or exceeding expectations, ranging from a high of about 43 percent in 4th grade to a low of 37 percent in 10th grade. Students did worse in math: 29 percent of Colorado students hit those marks, ranging from about 19 percent in 8th grade to 37 percent in 3rd grade.
20 largest districts’ 4th grade English scores
20 largest districts’ 4th grade math scores
DPS achievement gaps persist
The state’s largest school district, Denver Public Schools, continued its longstanding pattern of lagging behind state averages. But according to DPS’s analysis, the gap is smaller than with previous tests.
On the English language arts tests, which combine reading and writing, 33 percent of Denver students met or exceeded expectations, ranging from a low of 31 percent in 10th grade to a high of 36 percent in seventh grade.
Math scores in lower grades ranged from a high of 30 percent in 3rd grade to a low of 24 percent in 4th grade. Students who meet or exceed expectations on PARCC tests are considered to be on track to enter the next grade — or for high school students, ready for college or career. Overall, 25 percent of DPS students met or exceeded expectations in math.
Data Center | Find your school and district results in our database here.
“We were not surprised by the overall scores,” DPS superintendent Tom Boasberg said, explaining that the scores are in line with other tests that Denver students take.
Compared to the state average, Denver’s participation in PARCC was high — 92 percent of students took the tests.
DPS puts a high value on academic growth. Because PARCC is in year one, growth data, which measures how much students learned in a year compared to their academic peers, won’t be available until next year.
“The growth data is the most important data,” Boasberg said, because “it’s not about where students came in; it’s about what progress they made.”
While state officials are not finished crunching the test scores to show breakdowns by race and economic status at the district and school level, DPS has done some of that work.
Some groups of students, including students from middle- and higher-income families, did better on the new tests than their peers statewide, DPS says. DPS reached that conclusion by comparing the performance of district students who do not qualify for free- and reduced-priced lunch status with similar students in other districts.
However, achievement gaps within DPS persist. According to district data, white students continue to outperform black and Latino students on the tests. In fact, the difference between the scores in reading and writing grew even bigger than in the past.
20 largest districts’ 7th grade English scores
20 largest districts’ 7th grade math scores
Aurora: ‘We know we need to do better.’
Of the state’s top 20 districts, the troubled Aurora Public Schools district posted the lowest scores on the PARCC tests: 20 percent of students met or exceeded expectations in language arts, and just 12 percent met those marks in math.
With 18 schools considered failing by the state and the district’s accreditation hanging in the balance, the inner-ring suburban district has come under greater scrutiny as of late.
Superintendent Rico Munn, two years into the job, is urging patience as the district hopes to see results from reforms that include reallocating more than $10 million dollars to schools, changing how it recruits and retains principals, and creating a turnaround leadership team.
Munn called the district’s PARCC scores “unacceptable.”
“It is a representation of the work we have done before, and not necessarily a representation of the work we are capable of going forward,” Munn said.
“We know we need to do better,” he said. “We know we can do better. We are looking at this as the new baseline data set we can grow from.”
Munn said of the data on Aurora schools in turnaround status: “I didn’t learn anything. We know our schools that have the most significant challenges. This didn’t give me any new information about that.”
Other districts were not quite sure what to make of the data, in part because of high rates of students skipping the test. Colorado was considered a hotbed of the opt-out movement, and numbers released Friday showed the lowest participation figures in pockets of rural Colorado as well as wealthy suburban districts that usually post high scores on standardized tests.
(For more on PARCC participation rates at the district level, read this story.)
Overall, the state’s PARCC participation rate was 82 percent, an all-time low for Colorado state standardized tests. Under federal law, districts and schools are expected to have to test at least 95 percent of their students. A testing reform law passed last spring created a one-year timeout in the accountability system, so school and districts won’t face penalties for low participation.
Matt Reynolds, chief assessment and systems performance officer for the Douglas County School District, said that because there is just one year’s worth of PARCC data, it’s hard to draw any major conclusions based on performance or participation rates.
Dougco’s PARCC participation rate was about 70 percent, state data shows. Superintendent Liz Fagen said the high number of students opting out “does impact the (performance) data, no question.”
While Douglas County outperformed the state on PARCC as it has on other tests, Fagen said she couldn’t say whether the gap between the state and the district may have narrowed because of opt-outs. Dougco’s fourth grade English scores were 10 percentage points above the state average; in 10th grade English, the gap was 5.7 percentage points. In Dougco and other suburban anti-testing hotspots, opt-out rates are generally higher at schools that traditionally perform strongly on tests.
A number of district superintendents have been highly critical of the long lag time between the administration of the PARCC tests — which began last March — and the release of the scores. Officials with PARCC and its member states cited a process for setting performance levels on the new tests that stretched on for months, and promise a quicker return starting next year.
20 largest districts’ 10th grade English scores
Scores of districts on academic probation
These eight school districts are in jeopardy of losing their accreditation with the state if academic gains are not met in the next two years:
Concerns about participation, privacy
Friday’s release is another milestone in a years-long effort to develop and put in place stronger academic standards, and introduce computer-based tests to see how students measure up.
Unlike the state’s previous bubble-tests, PARCC tests developed for a coalition of states pledge to reward critical thinking and problem solving rather than rote memorization. The academic standards, which were adopted in 2010, are supposed to put students on track for going to college or starting work.
Yet PARCC and the Common Core State Standards have been heavily criticized by some as being misaligned and a federal overreach. Others claim too much time is devoted to the tests and that the results are used to penalize teachers.
Critics of the testing system are optimistic about potential changes under a new federal education law, which President Obama signed off on this week, that significantly shifts power over testing and accountability away from the federal government and to states.
The PARCC results largely mirror equally rigorous tests such as the National Assessment of Educational Progress and the Programme for International Student Assessment, said Joyce Zurkowski, the state education department’s assessments chief.
“We’re hearing that — frankly — as school districts look at their results they are not completely unexpected,” Zurkowski said.
Zurkowski said that while the data leaves some holes, she’s confident schools can find meaningful insight from their results.
“I think that participation is relevant for folks to be looking at,” she said. “But this is the starting point.”
She believes schools, districts and parents can compare results from schools that serve similar demographics and — at the least — tested a representative sample of students. Districts will deliver PARCC report cards to parents, with the timing up to districts.
“I think there is enough meaningful data to make a lot of comparisons,” she said. “Our comparisons are more limited because of the participation issues. If I’m a school, I’m not going to be interested in comparing myself to another that had only 10 percent participation.”
Because of increasing concerns about privacy, the Colorado Department of Education decided for the first time this year to obscure the results when there were fewer than three students in a category, whether that category was related to students who met expectations or students who didn’t. Officials said the fear was that someone with knowledge of a particular school or group of kids would be able to identify who those individual students were.
Graphics by Sarah Glen/Chalkbeat