Fixing Education’s Broken Windows: The Importance of Early Warning
A sixth-grader with a failing grade in math has only a one in five chance of graduating from high school six years later. This data from Robert Balfanz, the Johns Hopkins University researcher who revolutionized our understanding of the nation’s education crisis with his Promoting Power (or Balfanz) Index — and Lisa Herzog is absolutely sobering. And at the same time, the fact that we can actually identify students who are falling behind before they head into middle school (and even before they reach sixth grade) explains why we must use data in identifying and solving the broken windows that lead to so many kids falling into despair.
One of the dirty secrets in the battle over the reform of American public education is that so many of the issues that lead to kids failing in the classroom (and eventually, outside of it) can be easily identified long before it is too late. Thanks in part to the No Child Left Behind Act, the emergence of standardized and formative testing, and the early efforts of school reformers to improve data, researchers such as Balfanz can clearly identify when students fall off the path to high school and higher ed graduation. As Balfanz points out, 43 percent of potential dropouts can be identified by sixth grade, meaning that schools and districts can intensely intervene and help these kids before they reach high school. And while the conversations about dropouts tend to stem around the immediate issues that trigger students to finally drop out such as teen pregnancy, the reality is that the path to departing school before graduation is one that usually starts in elementary grades.
We now know that a sixth grader missing 36 or more days of school during the year has less than a one-in-five chance of graduating on time, and the same is true or a peer with discipline issues, while those students missing 18 days will also struggle to graduate. The data indicates that the students are struggling in their academic studies and have started tuning out of school; after all, no child wants to admit that they are illiterate or innumerate. Meanwhile the likelihood of a sixth-grade student with a failing grade in English graduating is even lower — just a one in eight shot. Essentially, these are signs that the kids have not mastered the basic skills needed to tackle harder reading and math subjects such as word problems. More importantly, those problems also manifest in tandem with truancy and other signs of dropping out. A sixth-grader missing 36 days of school, a failing mark for discipline, and failing math and English grades, will only have a one-in-10 shot of graduating on time.
Then there are indicators that come into view before sixth grade. For example, there is data on early childhood illiteracy, which can be measured through third and fourth grade reading tests. Twenty-three percent of third-graders who were functionally illiterate failed to graduate on time nine years later, according to an analysis of Peabody Individual Achievement Test Reading Recognition subtest data by the Annie E. Casey Foundation; one in six third grade students failing to read at proficient levels overall didn’t graduate on time nine years down the line. The data is culled from sample reports on some 4,000 students from the U.S. Department of Education’s National Longitudinal Survey of Youth 1979 — and thus not the best or most-reliable indicator of student achievement. But it does show the importance of identifying functional illiteracy during the first four years a child is attending school — and immediately providing struggling students intensive reading remediation before they reach fourth grade.
Thanks to tools such as the Dynamic Indicators of Basic Early Literacy Skills (or DIBELS) test, struggling students can be identified even before they reach first grade. There are also ways to help these students get on the right path before it is too late. Given that 40 percent of all kindergarten students can only learn to read if they are specifically taught syllables, words, letter sounds and spelling — and that boys, in particular, struggle because the area of their brains in which language and literacy is developed lags behind that of their female schoolmates, identifying these students and using new ways to help them improve their reading before fifth grade would keep them on the path to graduation. It would also help prevent the disciplinary issues that begin to crop up among students struggling with functional illiteracy by third grade (and help reduce the overuse of suspensions and expulsions that exacerbate the education and dropout crises).
Some districts are actually putting together their own early warning systems, albeit still on a small scale. New York City has taken some steps courtesy of Mayor Michael Bloomberg’s Interagency Task Force on Truancy, Chronic Absenteeism, and Student Engagement; but that effort so far targets just a smattering of the one million students who attend the nation’s largest district. A few other cities, notably the Diplomas Now project, which is working in Chicago and Philadelphia, are also developing early warning systems. States such as Indiana and Colorado have also done plenty of work on the early warning system front. But most traditional districts do little to identify children on the path to dropping out (much less offer any sort of intensive remediation or help dropouts return to high school and get on the path to college), while many states have done equally as little.
One reason lies with the problem of scale inherent in the traditional district model. Size can have many benefits, but not in improving the quality of education for students. As seen with the Los Angeles Unified School District, which evaluated just 40 percent of the Los Angeles Unified School District’s veteran teachers and 70 percent of new hires (who attain near-lifetime employment, and thus are far too difficult to dismiss, after two years on the job) during the 2009-2010 school year, districts already struggle in simply handling the human capital functions critical to improving student achievement. The fact that traditional districts struggle in the area of developing and managing data systems — with some systems storing data on FileMaker and Excel spreadsheets — also makes the development of early warning systems difficult to put together.
States haven’t helped in this regard. While statewide school data systems are becoming more robust, just three of them — Arkansas, Texas, and Florida — meet at least eight of the ten standards set by the Data Quality Campaign for being longitudinal and useful; and even those three states don’t provide access to data in a timely manner. Few states collect attendance data in any meaningful way, essentially providing little information on whether kids are attending school at all. Just 12 states collect attendance data daily (which students are actually in school), according to Balfanz’s Everyone Graduates Center, while a mere 11 states collect enrollment, attendance, and discipline data daily.
The fact that so much of school data remains compliance-oriented instead of being oriented toward accountability and usefulness in solving problems, is also an issue. That the measures aren’t useful also plays a part. Most states, for example, calculate attendance by dividing the total number of days missed by students by the total number of days they are supposed to attend (usually 180 days multiplied by enrollment); this hides the levels of truancy plaguing a school because it includes all unexcused absences, not just the set number of days under which a student is considered by law to be truant. So far, only California, Indiana, and Georgia provide breakdowns of levels of chronic truancy – and even those measures can be flawed because each of the states has their own definition of chronic truancy.
The federal government has proven helpful in the past in setting some standard for data through No Child’s Adequate Yearly Progress measures. But thanks to the Obama administration’s effort to allow states to waive the accountability provisions of the No Child Left Behind Act, that progress may be lost.. All but two of the 10 states granted waivers in the first go-round essentially ditched subgroup accountability by placing all poor and minority students into a super-subgroup that takes them off the radar, while other aspects of the waiver effort allow those states to let merely mediocre schools off the hook for student failure — and at the same time, denying reform-minded teachers and school leaders the data they need to make smart decisions.
Then there are the cultural realities within traditional districts. An early warning system involves using data in order to make decisions, and extensive collaboration within schools in order to put students back on the path to success; and thanks to No Child and other reforms, more teachers and school leaders are becoming savvy in using data. Yet there are still too many school leaders and teachers who don’t have the sophistication (or the desire to use data) needed to use do so properly; the fact that many school leaders still aren’t using Value-Added data in structuring teams of teachers who can address student needs (when they have that information available) makes clear the trouble of using early warning systems.
As for collaboration? Teaching remains largely an autonomous effort — and many veteran teachers like it that way; few instructors want to work together with colleagues in teams, much less working with guidance counselors and others on helping at-risk students succeed. This lack of teamwork has consequences. As Dropout Nation noted in its podcast profile of Harlem Link Charter School founder Steve Evangelista — who learned that a student he once taught as a teacher landed in New York City’s infamous Rykers Island jail — a struggling student loses contact with the one teacher that may have reached him, and further disengages from school. It also means that a teacher taking on a student with a long history of academic failure doesn’t know the particular issues facing that child and will have difficulty in getting her on the path to success.
Finally, there is the reality that far too many in education have low expectations for poor and minority kids. As Smith College professor Tina Wildhagen presented in her Teachers College report on the role of teacher expectations in student grading, African-American high school seniors were more-likely to get lower grades than their scores on 10th-grade math and reading standardized tests. From where some teachers and school leaders may sit, developing early warning systems to help struggling students would take time away from attending to those kids they deem worthy of their time and effort.
Certainly these challenges make developing early warning systems difficult. But it doesn’t make them impossible. There are charter schools and traditional districts and schools that are using data proactively in turning around the performance of struggling students. More importantly, developing systems to identify struggling students will not only help kids succeed, it can even help taxpayers save money in the long run — especially in stemming the number of dropouts on unemployment lines. And from a moral perspective, it is the right thing to do. There’s no way we can knowingly allow so many young men and women to continue into poverty and despair when we can identify their issues early on.
One critical step in making early warning systems more common starts at the state level with the development of more-robust longitudinal data systems that are geared in part toward identifying struggling students. Districts may need to join together on developing such systems in order to yield cost savings; this would be one of the few times that scale actually makes sense. This is also an area in which the private sector could do plenty of good; after all, companies can develop those early warning systems and then market them to the districts that need them. Because it makes far more sense to help kids succeed long before they reach third grade, formative diagnostic and summative standardized tests must be given as early as first grade just for diagnostic purposes.
The Obama administration could also take key steps towards this goal by ending its No Child waiver gambit — which will do far more harm to children than either the president or U.S. Secretary of Education Arne Duncan realize — and actually move to expand accountability and data; this includes developing a uniform chronic truancy rate of ten days of unexcused absence similar to what is already in place in Indiana. Expanding the Race to the Top initiative to include reform-minded districts that focus on developing early warning systems as part of their efforts would also help.
Ditching the traditional district model — and embracing the Hollywood Model of Education — would also help. But that is a long-term goal. Until then, districts will exist, and so we must do more to push districts to embrace the early warning system approach. One way lies with overhauling school funding itself; besides essentially taking over school funding and turning those dollars into vouchers that follow each student, states can also reward or punish district by the number of students they help improve achievement and turn around performance. This would encourage districts to use data in more-efficient ways. Those districts that are already making moves in this regard need to do more to encourage leaders on the ground in identifying student learning issues and in restructuring how teachers work (especially in the elementary grades, in which instructors are jack of all trades and specialists in none). Collaborative teams would certainly allow for teachers to focus on particular student needs, meaning that they will have to learn how to use data in more-sophisticated ways.
Finally, we must address the cultures of low expectations that make some teachers and leaders unwilling to actually help the students in their care reach potential. It means a whole revamp of how we recruit, train, evaluate and compensate teachers. Addressing those issues would do plenty toward giving our children the kinds of instructors and principals who make fixing the broken windows around them the top priority.