Test-bashing may be fashionable but two new polls show considerable support for testing by teachers, students and the public, writes Jill Barshay in Hechinger’s Education By The Numbers.

Graduation exams are backed by 77 percent of teachers and 86 percent of the public, according to *Teachers versus the Public*. “Accountability is one of the issues where the public and teachers agree,” said author Paul Peterson, a Harvard education professor. (Here’s more on the book.)

Ninety-four percent of students agree that tests are important for understanding what they are learning, according to a survey by Northwest Evaluation Association, a non-profit test designer.

In 2011, 60 percent of teachers said too much time was spent on test prep and test taking. In 2013, only 53 percent of teachers thought too much student time was devoted to testing.

“Formative assessment is shown to have the most positive impact on teaching and learning, yet it’s least understood and not widely practiced,” concludes the report. Only 29 percent of teachers correctly identified a definition of formative assessment. (Checking for understanding in order to modify teaching.) Most could not identify summative or diagnostic assessment either.

Tests are great for students if the students get the results in a timely manner. Then they can see where they need to improve.

That’s why kids enjoy Khan so much. The whole site is built around tests with instant feedback.

The problem with a lot of standardized tests is that kids get a number, but not a question by question breakdown that shows them what they need to learn.

I’ve never understood why standardized test couldn’t mean something like ‘at the end of the fractions unit, students can earn at least 7/10 on this set of problems’, with similar units for various skills. Similar standards for geography might be knowing the continents and oceans, or whatever is age appropriate. There wouldn’t be a giant ‘test of doom’ at the end of the year, but similar questions as teachers completed each unit, whenever and however they taught it. Most good teachers do these sorts of tests anyway, so I don’t see why their tests couldn’t contain a set of 5-10 ‘standard’ questions mixed in with whatever the teacher wanted to ask. Then teachers and students could be compared across classrooms, schools, and states, with minimal angst and stress.

I’m not sure that is a good idea. One of the advantages of a big late in the year test may be that it is more likely to assess actual long-term knowledge rather than short-term memorization.

If the teacher knows that there are “5-10 ‘standard’ questions” every unit that matter to the state, there is almost irresistible incentive to focus on those questions over and over (especially on the day before the test) so that most students can spit back the answer without thinking.

That may eventually lead to knowledge, or …

I see your point. I guess I was envisioning a test bank – you know that your students will have a fraction addition question, a fraction subtraction question, and an improper fraction question, for example. When you go to make out your test, you grab an envelope or download a set of 5 questions or whatever. I’m assuming that the teachers have some integrity and are actually trying to teach the students the concepts. I see the cheating scandals in the news, but the teachers that I know want the kids to learn. Of course, in my dreamworld the students repeat the unit until they can actually pass it. The next year, when they start the fractions unit, the teacher says ‘remember fractions?’ and gives a standard short pretest to see what the students remember so that they know where to start. This would asses the long-term memory issue.

Maybe I’m making this too complicated or assuming too much, but pre-and-post assessments containing standardized questions would seem to do what testing is meant to do, without dictating how and when content is taught. I can imagine a situation where teachers know that 3/4 + 2/8 is a question, so they focus on that particular problem, but that would be caught the next year so it would, in theory, not be a problem for long.

I think that the reason that this idea appeals to me is that I’ve taken and taught college courses that have common midterms or finals. Teachers were free to teach however they want, but they knew that their students would be tested on what the course claims to cover. With younger students, having these sorts of assessments before they move on and are too lost to catch up would help, and would also help avoid some of the issues that come up when the under-12 crowd has to sit for hours taking tests. Teachers and students would also have feedback quickly enough to actually use it, which, from my perspective, is the main point of tests in the first place.

I don’t think many teachers would deliberately cheat. It’s more the Feynman quote, “The first principle is that you must not fool yourself — and you are the easiest person to fool.” When the entire class moves together through the year, you desperately want to believe that most everyone is “getting it” well enough that you can move on to cover everything you are supposed to cover. And if those questions–forced upon you by the state after all!–say they are, well, believe it.

It would be different if they were part of a well-structured system of, “if you show you know this, and only if you show you know this, we will move you on to the next topic.”

Many math teachers of my acquaintance say they spend up to the first quarter of the year re-covering what the students had supposedly learned the year before.

I’ve had the job of teaching dosage and dilutions to pre-nursing students. There is nothing that you could tell me about how bad students are at arithmetic that would surprise me. And, I would LOVE a system in which students only move on if they understand the preceeding material, especially in math and reading at the early levels. Being able to pace math instruction has been one of the biggest advantages of homeschooling for us – we can skip over things that my kids can do, and I make up extra practice problems for things that they miss. Our curriclum has frequent short cumulative review tests, so if a skill starts to be forgotten, I add some practice problems for a few days.

Ah, if only all students could get that.