Using Data to Inform Education Policy

Secretary of Education Arne Duncan gave a speech on the importance of using data to inform education policy.

“There are already districts making exemplary use of data systems to let parents, teachers and administrators know how best to support their students,” he told a group of researchers at a conference sponsored by the Education Department’s Institute of Education Sciences.

How can teachers and principals use test results, attendance and other data to inform what happens in their classrooms?

ED Staff


  1. I am house-maker and I came to school in last October, by Online University, they offered to me a grant for 5000.00 never happen university sent me a Stanford loan for 12,0000. I was upset because supposed to be my option was a negative for more than 3,000 that money help me with my books. Now I have to pay 12000.00 and I am not afford to sent my daughter to the college and also they said you don’t qualify. What can I do I am a desperate mother I want my daughter going to the college. Why you don’t control those institutions. I work 1 hr. a day how I pay 200.00 a month when I afford the same amount per month.

  2. As a disability rights advocate, I have found special education data to be really useful in supporting my position when dealing with legislators and administrators. Some great sources of data are your state education department,, and the National Center for Special Education Research (

    I would like to see more data on outcomes for special ed students. I agree with the other posters that assessment data is not a reliable measure how well a student is prepared for success in the the real world. From my research, a student’s placement is a better predictor of future success than their test scores.

  3. This is a colossal waste of time and money. I diagnosed my daughter’s learning problems from IOWA tests 15 years ago.

    Data should be collected silently, automatically, in the background, as students do their routine assignments and tests online. That data should be analyzed in real time–mostly by computer algorithms–and virtual or personal assistance should be provided as needed. The way educators are collecting and using data is SO INEFFICIENT AND MARGINALLY EFFECTIVE, not to mention painful and expensive.

  4. Data can be quantitative and qualitative. Numbers only say so much, and reducing students and their education to numbers is problematic on so many levels. More qualitative data and research is needed to make better sense of the quantitative data we already have. Was that not common sense to begin with?

  5. Recently, data has become the panacea for educational reform. And the data being examined is from high stakes testing. In some schools, students spend more time preparing for, taking, and then reviewing test questions than they do actually learning.

    Concentration on individual students and universal design should be what we offer as professional development – not hours on evaluating “data” points. Incorporation of creativity, collaboration and cross-curricular studies should again become the goals of our schools.

  6. while i currently homeschool my dd did have to go through those tests once while i was sub-teaching in her district

    here is what i learned as a parent, a ‘teacher’ and a homeschool parent.

    1. districts with money throw A LOT of money at training the teacher how to teach the child to do well on the test…so who ever has the most money will do best on the test.

    2. teachers tell kids that if they don’t do well they will get left back, in trouble, etc… and the kids are overwhelmed by the tests

    3. the kids practice for the tests for months taking away from teaching time

    4. the tests test for the norm… if a child is outside of the norm (either side) they really don’t test well. assumptions of knowledge are made (like sports questions, tv questions, etc) and generalizations are expected to be understood (highly gifted kids don’t see what others do so these tests are very hard on them)

    5. the best way to check on a school is to have a warm body walk in and talk to the kids or sub-teach (without the teacher preparing the kids).. you would be surprised how quickly i could tell what kids/teachers were really doing when i was in the class with the kids.

    testing doesn’t work … it’s time to find a new way to supervise schools..

    oh, a national curriculum would be helpful so when kids move from state to state or district to district they wouldn’t have to catch up or wait.


  7. Some data doesn’t tell the true story.

    In Arkansas, it is possible to get Praxis score of 185 in mathematics and be denied opportunity to teach mathematics due to a score of 143 on a Mathematical Proofs and Models Exam. When the cutoff is 144 a teacher who can really teach math is screened out by the “data.” Rather than offering that person a chance to teach math with students who are not making the grade, that person falls through the cracks and ends up idle for the entire school year. That makes no sense.

    Data should be one consideration, not the entire deciding factor.

    The data sustains the “widget effect” and people who could be inspiring our students in math and science are instead ignored by a data-happy structure.

    Schools and administrators need to take a more proactive role in identifying teachers, substitutes and applicants who could be making a real difference in education but are on the sidelines due to politics, application procedures and data at the exclusion of common sense.

  8. You want to use data to inform instruction?

    Fine. Here are the essentials:

    1. No “high-stakes” data. Data should never be a weapon to threaten schools, teachers, or students with. That’s not an appropriate use.

    2. No data just for the sake of having some “data” to look at. Let’s not waste time collecting and analyzing weak or non-essential data. Only HIGH QUALITY data deserves our time and energy.

    3. Who should decide what constitutes “high quality” data? The people collecting and using the data: teachers.

    4. If we’re going to spend the time collecting and analyzing, we’re going to need time to plan and apply how to address the needs indicated by the data. Don’t underfund, or partially fund; that’s a waste of our time and $$. Either fully fund a system that pays for all of that analyzing and planning in addition to the regular instructional day, or let it go.

    5. And finally, let’s be honest about what really affects standardized test scores, if they are to be part of the “data.” We all know that the biggest predictor of standardized scores is not what the teacher does; it’s parent SES. Are you ready to address the issues of poverty and the undervaluing of education in the U.S. culture as a whole? If not, then the data is not a very good tool.

    Are you ready to fund class sizes and teacher time outside of instructional hours to make it possible to focus more intensely on the individual strengths and weaknesses of every student, or are you going to expect that data to somehow create miracles in the current factory/business model we operate in?

    The data is not god, and more data does not always result in more effective teaching or learning.

  9. I worked in one of those districts for 20+ years. The methods of data collection and data use have improved greatly during that time.

    I will begin by saying that I agree data is important. Used wisely, it can inform practice and can be a diagnostic tool when encountering challenges to student learning…our goal for every child. However if, rather than being seen as one tool among many, data becomes the number one driver of every aspect of what educators do, then I think we are choosing to use data unwisely. And we do so at our peril.

    THAT, in a nutshell, is the problem with the last years under NCLB…data has become the ‘be all’ and ‘end all’ in education and it has been applied unfairly to students, teachers and probably even some principals.

    Principals, as at-will employees, know they will be replaced if their school does not meet AYP expectations. Their goal, rightly, is to create the best school for their students possible. Data is a tool that can help them. They can see which students need support and then gear programs to provide that support. They can see which students are doing well and reward them. They can, similarly, view data by and about a teacher and use it to reward, support or punish a teacher.

    Teachers can also use data to diagnose and support student instruction. If a student is having difficulty (in math, for example) test data will often help find the problem so the teacher can work with the student to solve it. There are two thoughts I have, though, about data for teachers (as my district relied on it heavily). One, most teachers don’t need the data to know what the problem is, although the data often confirms a problem and documents it for administration. Two, teachers are given TOO MUCH DATA to ever have time in the day to use it effectively. Teachers should be given succinct and specific data about students…detail should only be given when asked for to diagnose a particular problem.

    As to ‘other data’ besides test and attendance data…I am concerned about privacy issues. I once had an administrator who required every teacher in my school to take their student class list and add anecdotal remarks which was put into a database for the school. I think that violated privacy rights. (The kinds of information gathered were things that parents might have shared at conference time.) It was ordered to be done the last week of school and teachers were prevented from checking out until it was turned in. I believe that crossed a line.

    So, as we incorporate data as a tool of public education, I hope we will use it fairly for all…students, teachers, parents and principals.

Comments are closed.