How Much Do We Test?

Day of the Dead
San Francisco's Day of the Dead procession - photo sent by Ellen Schneider, her comment on what is happening in public education

The results of our survey are beginning to come in. In hopes of encouraging more superintendents and school boards to provide data, I offer this early report, with the caveat that it is based on information from a handful of the nation’s 14,000 school districts.

The average number of mandated, machine-scored standardized tests administered in a district in a year: 22.

Of those, all but one are reported to be mandated by the state, not the district. (While there are no federally mandated tests in public education, the federal No Child Left Behind Act does require that students in grades 3-8 be tested in reading and math. {{1}})

Eleanor Chute of the Pittsburgh Post-Gazette cautions that some mandated standardized tests, such as DIBELS, are not machine-scored, which means the reported number may be higher. In September Ms. Chute reported on the steps that the Pittsburgh School Board was taking to reduce ‘excessive’ testing. Those eye-opening numbers deserve your attention:

The biggest reductions are planned in grades 3, 4 and 5 where the number of periods spent in testing are to decline from 85.5 periods to 41.5 periods. After school board member Sherry Hazuda was told one period equals 45 minutes, she said, “No wonder people are complaining when you see it like that.”
“This has a great impact on the amount of time our children have to learn,” said board member Carolyn Krug, who said it will reduce stress as well.
In kindergarten, the number of periods spent testing is to be cut from 13 to 11. In grade 2, the reduction is from 45 periods to 26. {{2}}

Districts report that the average lag time between when students take a test and when the results become available is nearly two months.

This confirms the conventional wisdom. However, the situation may be worse than it appears because so-called “end of year” tests are often given well before the school year actually ends. If those tests are administered in mid-May {{3}} and school ends in late June, then the results–which arrive during the summer when teachers are off–don’t come into play for four months, and only after the kids have moved to new classrooms and new teachers. Here’s where the journalist’s question, “Who benefits?”, is worth asking.

We asked how many mandated tests a student would take if he or she went through the system from kindergarten to graduation. The average: just over 37 tests in 13 years…

Is that a lot? Kids generally aren’t subject to mandated, machine-scored standardized tests in kindergarten, first or second grade. The pattern I am most familiar with is testing in grades 3-8, grade 10 and grade 12, and only in English Language Arts and Math.

Districts report that, on average, testing occurs in about 40 of the 180-day school calendar.

If you’re a percentage person, that means that on 22% of the school days, some children somewhere in the district are taking one of these tests {{4}}. It does not mean that schools come to a standstill on those 40 days.

How about ‘test prep’ time? Here I fear that the numbers reported so far may not be reliable. One district said its teachers spend ZERO hours preparing kids for tests, while another reported its teachers spend 100 hours.

So far no school district has reported disciplining, firing or giving bonuses to teachers based on student test scores.

Many states now require using student performance as a piece of teacher evaluation. How much scores should count is what is being debated. Judging teachers by test scores is only going to increase down the road, because the Administration’s “Race to the Top” program requires it.

No district reports that cheating by students or adults has been a problem.

We need more hard data.




Number of Students in District _______
Name of District ______ (this will be masked in survey results)

Students in our district will take ____ standardized, machine-scored tests during the course of the year. (We are seeking the number of different tests given. So, if a particular test is administered to three grade levels, that counts as three tests, not one.)
Of these tests, ___ were selected by the district, and ____ are required by the state.
A student who attends our district from Kindergarten through graduation will take ____ standardized tests during his or her time in school.
There are ____ days on our 180-day calendar when a standardized test is NOT being administered.
Some test results determine a child’s promotion, while others are used to evaluate schools and teachers. Is this the right approach? How useful is this information? Should we test only a carefully drawn sample of students when we want to evaluate schools, instead of testing every student?
Right now we test all students in only ____ subjects. However, if we are going to evaluate all teachers according to their students’ scores, we will need to test in more subjects, including art, science, social studies and music.
There is a ____ month lag time between the administration of a test and the delivery of the results.
Our school district spends $_____ per year to buy and score standardized tests. (These out-of-pocket costs do not include the labor costs.) This amounts to less than _____ % of our annual school budget.
We devote _____ hours per year to testing, out of approximately 900-1000 hours of actual class time during the 180-day school year.
In addition, many teachers devote an estimated ____ hours to what is known as ‘test prep.’ The numbers vary from teacher to teacher and school to school and are difficult to pinpoint.
Last year we removed ___ teachers because their students’ scores did not go up sufficiently.
Last year we distributed $$ _____ in cash awards to teachers whose students’ scores went up well beyond the expected level.
Last year we investigated ___ incidents of alleged cheating on standardized tests, ___ by students and ___ by teachers and other administrators.


[[1]]1. Thanks to Diane Ravitch for calling my attention to my initial inelegant phrasing.  The Feds don’t say which tests to use, but they do require testing.[[1]]

[[2]]2. Her story on the extent of testing in Pittsburgh received (well-deserved) national attention. [[2]]

[[3]]3. For example, Pennsylvania’s “Keystone” exam, the PSSA (for Pennsylvania State System of Assessment) is given in mid-May.  Pittsburgh is moving up the opening of school to try to get more of the curriculum covered before the kids are tested on it—what a concept–which means that the lives of children and families are being reordered to accommodate the machine.  Why not move the test date to later in the school year?  Because then there wouldn’t be enough time for the machines to process the answer sheets!  [[3]]

[[4]]4. Please bear in mind that the data presented here does not include teacher-made tests (which are generally recognized to be the most effective way of measuring student knowledge). [[4]]

15 thoughts on “How Much Do We Test?

  1. Footnote 4 parentheses is incorrect. There are no data to show that teacher-made tests are the “most effective” ways to measure student knowledge. Indeed, study after study shows that local tests are far less valid and reliable than professionally-designed tests – as we should expect.

    In our audits of school and district tests we find error after error of a most basic kind in test construction and lack of rigor.

    By my math, some conclusions here don’t add up. if the average student takes 2 tests per year (37 divided by 13) that hardly adds up 22% of days for EACH student.


    • Grant,
      I relied on memory….apparently a mistake. But then why is it that student grades are the best predictor of future performance, since those grades are by and large the compilation of performances on teacher-created tests and requirements?

      Re the 22%, that’s the reporting districts’ estimate of how many days when tests are being given somewhere to some kids, not when testing takes over.


    • Hi Grant,
      To your point about preferencing local, teacher-designed tests over “professionally”-designed tests. Remember Elinor Ostrom who won the Nobel prize in 2009 in “Practical Economics” – a practical economic theory better known as “the science of collaboration?” She would totally agree. Retiring the tired “tragedy of the commons” theory (in which the people inevitably spoil a common good out of greed, laziness, and neglect), she posits that actually it is only people with local control of a common good – in education, that would be our common interest in educating fertile minds – that can realize its potential … because locals – the classroom teachers – have a direct (not theoretical, political, financial) knowledge of the actual 30 or so children and what they need. Ostrom proved that it is well-meaning but misplaced intervention from far-flung authorities that keeps the good from being realized.

      “What we have ignored is what citizens can do and the importance of real involvement of the people involved – versus just having somebody in Washington make a rule.” – Elinor Ostrom

      Oh, and … those teacher-designed tests are the true professionally-designed tests – those others are commercially-designed tests. Everyone knows that fresh is better than canned. 🙂


  2. John,
    There seems to be a discrepancy between the number of tests a student is likely to take over 13 years (37) and the number of periods devoted to testing in a single grade or cluster of grades (26 periods in second grade, 41.5 periods in grades 3-5 in Pittsburgh Public Schools after significantly reducing the amount of testing). Do you know what lies behind this?


    • I suspect that some districts are under-reporting. I cite Eleanor Chute’s reporting because she dug deeply into the data. My hope is that readers will press their school districts for information.


  3. “While there are no federally mandated tests in public education, the federal No Child Left Behind Act does require that students in grades 3-8 be tested in reading and math.”

    So it is literally true that the Federal Government does not mandate specific tests, it does mandate testing. Splitting hairs? Yes, and badly…

    I am getting a lot of (well deserved) grief for my phrasing.


  4. Hi thanks for keeping it up day in day out. I appreciate it.

    I am a contributing founder of the first public Waldorf school in Los Angeles and the former director of development of the Association of Waldorf schools of North America.
    As you probably know, private Waldorf schools do not do any form of standardized testing. Despite this, the only perfect score on the exit exam in England was a Waldorf student. Another Waldorf graduate, from Germany, just won the Nobel prize in medicine. So lack of testing obviously doesn’t hinder exceptional achievement in school and in life. Finland also. Like Waldorf and other progressive pedagogies worldwide, begins formal reading at 7. Rather than a regime of testing, they apply cognitive science, developmental psychology, and healthy biorhythmic scheduling.

    Okay that was my soapbox.

    The reason that I’m actually writing is to question your assertion that there are no federally mandated standardized tests. While that may be technically true in fact it is not actually true. Correct me if I’m wrong. The two consortia, smarter balanced and PARCC get the funds from the federal government and with those funds they pay Pearson and McGraw-Hill et al – the education industrial complex – to develop the tests. Without federal money those tests would not be developed and if those tests were not develop those tests could not be given.

    Furthermore, by bribing states to sign on to the common core, the federal government, in the form of race to the top, is signaling it’s preference for A single unified standardized approach to knowledge.

    People who say, it’s the states, not the federal government, are confusing and apples for an apple.

    The only appropriate place to put our trust is in teachers and teaching. What might happen if the consortia were to put that federal money into building a community of responsibility around each single public school in our country? Farm to table breakfast lunch and dinner. On call healthcare and every school. Psychological support for every child and every family on call at school. I am arch instructors sure that every child can learn A musical instrument, singing, drawing painting dancing and gymnastics. Take the asphalt off the schoolyards and replace it with meadows, “loose parts”, an adventure playground, ponds.

    Wouldn’t it be better for our tax dollars that the federal government distributes to profit children and youth directly then to directly profit Pearson, McGraw-Hill?

    That’s all!

    Studio City, CA


  5. Joan
    The picture is going to change dramatically this Spring, but we asked districts to report on current data, which meant last year’s numbers.
    As you have no doubt figured out, my hope is for thousands of conversations in cities and towns, in political gatherings and around dinner tables, about assessment, testing, the purposes of education, et cetera.
    My belief is that good solid information is the best starting place, and I am so grateful to you for adding your perspective.
    PS: We filmed in a Waldorf school in California this spring. What a wonderful school….so full of joy and positive energy.


  6. Are students being overtested is a good question to ask of schools, districts and states. There should be more transparency and informed discussions about reducing testing. To make better decisions, more information about testing should be documented. Achieve created the Student Assessment Inventory Tool to gauge just how much testing a student sees in a year. School districts in Connecticut piloted the tool earlier this year and now it’s ready to be used by any disctrict across the country. We hope the result of districts using the tool will be tfewer and better tests.


  7. I’m an “external representative” on the LAUSD Early Childhood Education and Parent Engagement Ad Hoc Committee. One of the presentations regarded the reauthorization of Early Childhood Education grants. Our legislative rep mentioned 10% of the funding would go towards “Quality Improvements”, some which involve “assessing” our 0-4 pre-schoolers. I asked if these “assessments” involved collecting data. It seems that part of the justification of funding is to place personal information in a database controlled by an outside private entity. When I asked if there could be a system of identifying by numbers instead of the more personal data, I was told that this information would “follow” the child up into K-12 to “help” with academic need. Why can’t this be done without divulging so much personal info? Why are we getting into testing them at such a young age? It’s out of control and invasive at this point. The family of a child with an IEP is supposed to expect confidentiality, but all that seems to go out the window. I’m sick of “grants” anyway. Fully fund education and quit parsing it out to a few schools here and a few schools there. They all need funding.

    Those reformers who want all this testing and Common Core data send their children to private schools with small class size, enrichment and – oh, yeah – less testing. If they’d quit hiding taxable income in offshore bank accounts and pay their fair share, we wouldn’t need this ridiculous grant system. It’s a joke when you see itemized accounts of how much actually goes to each school in the form of a “reward”. Peanuts. We’re capable of much better, but I guess it’s better spent on war and prison.

    Sorry I got a little off-topic, but it is sad to see school districts scrambling for funds through the grant process that they may or may not receive. How can anyone plan a budget?


  8. Such misleading measures, all underestmates of standardized-testing burden. DCPS has for years not only tested and test-prepped students for the late-Spring ‘big one’; all students were also subject to four tests during the school year, all sold to the school system as so aimilar to the ‘big one’ as to be predictive and diagnostic, enabling the teacher, school, and system to target instruction.
    I find no recognition of the burden of the pre- tests or test-prep for any of them, either above, or in comparison among 9-11 tests coming on line now. As though any or all of the practice and pretesting, will be going away.
    Imagine a goal of student-body physical fitness. Suppose fitness were measured by final standing of the football team. Who would measure the time-burden for physical fitness by the minutes played in the championship game –48-60 — and lgnore all the season and pre-season games and all the practice and weight-lifting time


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s