Could you imagine a city that didn’t closely track the value of every house to assess property taxes? This data is ubiquitous and drives both individual decisions (like home buying) and policy decisions around budgeting, planning, zoning, and transit. This information is available to the public and policymakers nearly immediately and is just a google search away.
Data that should be informing policy and communities is not as readily available as housing valuations. For example, information about police misconduct, substance abuse and treatment, and insurance rates, amongst others, are critical to understanding community health but are often hidden from the public.
In contrast, Colorado has long been a national leader in making information about its education system transparent. Yet, recently Colorado state education officials have made a series of decisions that make it impossible to know how our most challenged students are doing in school.
Colorado has spent twenty years leading the way on data transparency and school accountability. We were one of the first states to adopt state standards, a robust state assessment tied to those standards, develop an academic growth measure, balance student growth with measures of proficiency, and to publicly provide student performance data through web-based tools. Taken together these developments enabled policy makers, educators, community stakeholders, and researchers to better understand how students (and, critically, different groups of students) were performing in any Colorado school.
Policy, practice and bipartisan leadership at all levels of Colorado government have encouraged education data transparency so that we might better understand what is working and not. It is the backbone of any effort to improve public education.
Yet today, as students across the state are wrapped up CMAS PARCC assessments for the third year in a row, the data and insights Colorado offered from these annual measurements of whether or not Colorado students are on track to be college or career ready are woefully opaque.
It was a full 14 months after the administration of the 2015 CMAS PARCC assessment that the State released information about whether different groups of students met grade level expectations. And even then the vast majority of that data was not reported. For example, over 70 percent of the district and school level data fields about students by family income were masked.
And last year’s test results? While the State has released average scores on the test by some student groups (though not all), disaggregated data about whether or not students are reaching grade-level expectations has not been released.
We have previously mentioned the limitations of releasing average scores, but it is worth a deeper look. The only groups of students for which the state has released average scores are those groups that are counted specifically in the accountability system. That means there is information about average performance of students eligible for free or reduced price lunch, but not students who do not qualify for free or reduced price lunch. There is information about average scores from “minority students” (CDE’s language, not mine), but not white students, and not by “minority” races or ethnicities. The achievement gaps by race and income are arguably the most vexing issue facing society and our education system yet we cannot know how large these gaps are or where these gaps are improving or growing because this most basic achievement data is unavailable.
A few factors spurred the twenty-year leap backwards in Colorado’s data transparency. The transition to new assessments, though Colorado had been administering a statewide standardized test since 1997 opened the door for more regressive data sharing policies. At the same time, there has been a growing backlash against school accountability, despite the fact that Colorado has not forced any schools or districts to make radical changes since the ham handed take over of Cole middle school over a decade ago. There have also been rising concerns about data privacy particularly as more technology and tools from different vendors appear in classrooms. In response to each of these pressures, the state has hidden more data from the public, leaving us worse off than we were twenty years ago.
This has come at a real cost. I knew more about how low-income students were doing in a district like Aurora, Roaring Fork or Denver back in 1998 than I do now. It is absurd that despite better testing and data infrastructure, we have less information. The information exists but it is hidden behind a series of rules that claim to protect students. The irony (or outrage) is that it is hurting the thousands of students who are graduating from Colorado schools with none of the skills or knowledge necessary for a living wage or college.
Colorado’s own ESSA state improvement plan says “As a result of these practices, Colorado has one of the most conservative public reporting approaches in the country. The chances for individual student assessment performance level information to be calculated or inferred in Colorado has dropped dramatically.” The plan does not expand on either the magnitude of the challenges or the burgeoning solutions in our state.
Unlike just about any other public education problem, this is an easy problem to fix. The State Board of Education and/or state legislature could pass a rule and/or law that would require that the state issue disaggregated school level data with an “n-size” of 16 (as is in our ESSA plan) at the same time any achievement data is publicly shared. This has been Colorado’s practice in the past (that has resulted in ZERO data breaches in terms of individual students), and is common practice is most states. The current practice also appears to be violating federal law under our NCLB waiver and what is understood under ESSA. It’s time for Colorado to get back on track to improve public education, start counting all students again, and move forward with improving our schools. Our communities and economy cannot afford to go backwards decades.