The True Cost of Testing

Next week my school is giving the state-mandated testing to our freshmen, sophomores and juniors. Colorado, like many states, has made some changes this year, and they are definitely an improvement. The overall amount of testing is decreased somewhat, and sophomores will no longer be taking the PARCC exam but instead will be taking the PSAT. While I'm not a fan of PSAT and SAT either, they are at least somewhat useful to some students.

When the news was announced in December, the focus was primarily on three things: the reduced amount of testing, the switch from the ACT to the SAT, and the timing of the announcement. What was missing was much discussion of the merits of testing in general and the cost (both direct and indirect) of the testing. For this post, I'm going to just focus on cost.

I've so far been unable to find anywhere on the Colorado Department of Education site the cost to the state of these state-mandated tests. (I'm sure it's there somewhere, I just haven't found it yet.)  This article seems to indicate that the current ACT for juniors costs about $2.1 million a year, and they are budgeting $1.8 million for the sophomore exam, plus an additional $432,000 for juniors who want to take the writing portion. The cost for PARCC/CMAS (9th graders and 11th graders in science) is harder to figure out, but this article from about a year ago shows the state will pay Pearson about $27 million for PARCC and CMAS. If we assume that the high school portion of that is about 10% (very rough estimate), then add in another $2.7 million. So, somewhere around $7 million in direct costs to the state. That figure, of course, doesn't include the indirect costs of staffing, materials, time, etc., nor does it include the same types of indirect costs to school districts.

But even what little focus there has been on that really quite large sum of money then ignores the opportunity costs that are ultimately paid by school districts (and kids). Let's look at my building as one example. Next week we will spend three days on state-mandated testing. While we run an abbreviated schedule in the afternoons, most folks will acknowledge that the classes held during that time are not optimal for learning. The students who took the tests in the morning are tired, and some students who do not have to take the tests in any given morning choose not to come to school just for the classes in the afternoon (definite surge in our absentee rate). Given all that, many teachers make the reasonable decision to limit their instruction during this time to less critical matters. Not that we don't try to make the time worthwhile, we definitely do, but it's tough to try to reach the same level of learning as in a typical day. So, for me, I consider those three days pretty much lost for instruction. (If you disagree, you can pro-rate the numbers I'm about to share accordingly.)

So what does this "cost"? Well, according to district budget documents, we spend $9597 per student per year (that's including federal, state, and local funds; the amount directly from the state is less than that). We have about 175 days of "instruction" (theoretically), so $9597 divided by 175 works out to just under $55 per day per student. Since we have roughly 2150 students at Arapahoe, a single day of instruction costs roughly $118,000. That means that for each and every day of instruction we choose to "give up," we are "forfeiting" that money. So each day of testing is costing us $118,000, or roughly $354,000 for the three days of state-mandated testing. (Keep in mind that does not include the pro-rated cost of the $7 million the state is directly paying, or the indirect costs to the state and especially the districts that I'm sure adds several million more.)

But, sticking to just the lost instructional time, we're now at $350,000 (just at my school). But there's more, of course. We currently choose to take a day of instruction in the fall to give the PSAT to all Juniors. (Yes, despite the fact that we're now going to be giving it to all sophomores in April, we are still going to turn around and give it to all of them again in October when they're Juniors. Why? National Merit.) So that choice means we're deciding to spend another $118,000 on testing. We're now up to $472,000 (just at my school).

But, of course, there's still more if you don't limit it to state-mandated testing. What other types of testing do we have at my school? Well, we give MAP testing in language arts and mathematics to our students. That's a bit harder to quantify in terms of cost, since we don't devote parts of entire school days to it. Instead, students are tested in their language arts and math classes (twice a year in 9th and 10th grade, just once a year in 11th I think). Making a very rough estimate again, I'll say that equates to about half a day per year per student, so $59,000. We're now up to $531,000 (just at my school).

We also have many students who take AP exams at the end of the school year. When students take an AP exam, they not only miss the 3 or so hours they are writing the exam, but they often miss the entire school day as they are pretty exhausted. There are certainly many arguments in favor of the usefulness, importance and value of AP Exams, but there are also arguments against. No matter which side you fall on, certainly those days are not available for instruction for those students. (And even for those students who are not taking an AP exam, teachers adjust what they are doing in class because so many students are missing due to the AP exams). I don't really have the data to completely quantify this, but our students write close to 900 AP exams in a given year, so if we take 900 times $55 per day per student, that adds $49,500. We're now up to $580,500 (just at my school).

My school also requires final exams each semester. We devote four days each semester exclusively to final exams, so eight days throughout the school year. While there are certainly some folks who will argue that final exams are useful, necessary and important, there are also arguments that they are not. Whichever you believe, they certainly don't much resemble instruction. So eight days times $118,000 adds another $944,000, so we're up to $1,524,500 (just at my school). Since many teachers also take at least one day to review for the final exams, you could perhaps add in more here (although that review has some instructional value, so I'll leave that out for now).

Then you add in the individual tests that teachers give. This is even murkier territory, since I do believe assessment - when it is done well - is very valuable, and how teachers give these assessments varies tremendously. But certainly there are a fair number of teachers who give "unit" tests multiple times a year that take an entire class period. That's time that is no longer available for instruction, so there is an opportunity cost associated with it. Let's make a conservative estimate and say that each students loses 4 days a year cumulatively to these tests. That adds another $472,000, so we're up to $1,996,500.

Now, that's waaaaay too many digits of precision, so let's just say $2 million as a rough estimate. We spend $2 million a year on testing . . . just at my school.

Two. Million. Dollars.

We can debate my estimates and I'll freely admit that I'm just ballparking all of this, but at least it gives us a place to start the discussion. If we ignore the $7 million the state spends in direct costs, and if we ignore the additional millions the state and school districts spend in indirect costs, and just focus on what my high school spends on testing each year, $2 million is a good number to work with.

That's the cost of testing.

Yet even that isn't the true cost of testing, or at least not the total true cost of testing. My daughter is a sophomore this year, and in her language arts and math classes they have been doing some practice PSAT items. They are not spending a lot of time on this, but they are spending some. And just because they are practicing for the PSAT doesn't automatically make it a poor use of time, the skills they are practicing may (or may not) be valuable.

But I think we have to acknowledge that in addition to the actual time spent testing, we are impacting what we do in our schools. Even if you believe those practice items are valuable, keep in mind that those items change each year as the tests change. We used to do CSAP practice items, then TCAP, then PARCC, then ACT, and now PSAT and SAT. While those are certainly related, each time the test changes we change the prep we do. I think it's awfully hard to argue the high ground here about how valuable these items are when they keep changing based on which test we're giving.

And it doesn't just influence those test prep items. We change what we do in our classes based on these tests. From major changes like adjusting the entire curriculum, to more minor changes like materials selection and the emphasis we place on different topics within that curriculum, the current test ends up driving a lot of what we do (even if we don't want to admit it). Again, that doesn't necessarily mean that the things we are doing are bad, but I think we need to be honest and acknowledge why we are doing them. The question we need to ask is what would we choose to do with our students in the absence of those tests? Instead of trying to do things better, we should do better things.

It's not just the time (and dollars) spent on actual testing, it's the impact on the culture of learning in our schools. It's the message we send to students (and teachers) of what we value and who we are serving.

That's the true cost of testing.