Dear Data, Please Make Yourself More Useful
Sincerely, teachers and students
As surely as the trees bud in spring, night turns to day, and the Kardashians provide grist for the tabloids, another education practice—the use of education data—is turning ugly. Factions are setting up camp at two extremes: one for those who believe data is the Holy Grail, and the other for those who shun it.
Meanwhile, our students are counting on us to help them learn and be successful. Consequently, we believe there is a way to acknowledge that both sides have valid concerns, while applying a "usefulness" standard to make sure we're collecting information that actually can be drawn upon to change schools for the better.
While mountains of data exist, there is little that busy people can use to make good decisions. Educators are natural cynics, and their daily interactions with students are often dramatic proof of each student's qualities and the vagaries of growing up. The fixed and standardized ways that data are reported often do not strike educators as relevant or useful. But by focusing on students and the value that data can provide to better understand each one, we change the dynamic, win over teachers, and improve student learning,
Just as electronic health records provide doctors with access to a patient's full medical history and reminders about particular health issues, so education data can provide teachers with insights into a student's learning history and unique needs.
Here are some guidelines for meeting the standard for useful data:
Engage teachers and decisionmakers in the design of the tools used to collect data. According to a recent article in Governing magazine: "Forty states provide school principals with student longitudinal data" that follows student progress from grade to grade, while only 28 do so for teachers. And, Governing says, "40 states offer feedback or growth reports to teachers based on student-performance data." But, we'd add, too few ask whether the data included are what teachers want and need. Asking those who perform the work to provide input in the design of the data-collection and -reporting tools they will use makes abundant sense.
Unfortunately, teachers haven't been as involved as they should be in the development of education data systems, and it shows. Instead of an array of indicators that teachers can use to make midcourse corrections and revised lesson plans that acknowledge their students' needs while learning is in full swing, the emphasis is on summative test-score results, which measure learning at the end of a course of study.
Create regular opportunities to huddle around the data. Again, according to Governing magazine, only eight states require teachers and principals to be "data literate." In addition to setting aside time for training, statewide longitudinal-data systems should create regularly scheduled opportunities for teachers to gather and strategize about particular students who are struggling. Data systems like the California Partnership for Achieving Student Success, or Cal-PASS, which has collected years of school transcript information for more than 40 million students, show what is possible when educators move from blunt, end-of-year test scores to detailed and timely student-performance information.
For example, San Diego-area high school teachers and college faculty members learned through careful review of Cal-PASS data that, counter to conventional wisdom, students who took English courses through 12th grade were just as unprepared for college as students who stopped taking English courses after 10th grade. Working collaboratively via professional learning councils, San Diego educators determined that an almost-exclusive focus on literature in high school wasn't giving students an opportunity to develop the writing and analytical skills they needed for college and careers. Subsequently, San Diego high schools began teaching more writing and critical thinking.
By responding to this important indicator, teachers in the English Curriculum Alignment Project, or ECAP, kept 86 percent of their students on course to successfully complete college-level English. In contrast, only 24 percent of students placed in the lowest level of English remedial courses in California colleges ever make it out. The collaboration is an extraordinarily uncontroversial effort which teachers and administrators universally support.
Tailor reports to your audience. There are so many stakeholders interested in how schools are performing, but they often want different things at different points in time. While teachers focus on their classes or specific students, superintendents may examine the impact of a new curriculum or teachers hired from a specific college. At the same time, parents are more likely to look at school and teacher-level performance. Some districts have learned the hard way the limitations of what data can and cannot show. That doesn't mean the underlying data was useless; it just wasn't the right tool for the job (akin to measuring air temperature with a stethoscope).
"Useful" means many things and has many audiences. Currently, the data that school systems collect and report to states is too often limited to only what is required under the federal No Child Left Behind Act: standardized-test results for reading and math in grades 3-8; science- and writing-test scores in at least one grade at the elementary, middle, and high school levels; and graduation rates. While valuable, this information will be vastly improved with data that teachers can use to tailor lessons to students, such as which courses students took in prior years and the grades they received, and the students' writing samples, diagnostic test results, and participation in tutoring programs.
Continuously hone validity and accuracy. The exclusive focus on summative tests and "accountability" often viewed as punitive and unfair risks a serious crisis of confidence in the power of education data. As some school districts have shown, student test scores alone are not valid measures of teacher performance. Instead of tunnel vision focused on narrow test results, statewide longitudinal-data systems have the opportunity to become highly developed instrument panels that guide teaching with a host of information about students, not just test scores. Further, educators engaged in using Cal-PASS report that the daily practice of using data not only improves teaching effectiveness, but also improves the data. The more educators study it, the more they understand and can perfect what is being measured.
After hundreds of millions of dollars and years of tinkering, the time is nigh for education data systems to make themselves much more useful. Just as electronic health records and disease registries are fueling greater discoveries and personalized patient care, education data must become a necessity of teaching. We wouldn't think of cutting back on data in medicine, because it is endlessly useful. The San Diego-area students now excelling in college-level English—a key gateway for success in career and life—would say the same about their teachers' use of education data.
Vol. 31, Issue 32, Pages 22-23Published in Print: May 23, 2012, as Dear Data, Please Make Yourself More Useful