Race to new tests

Competition has opened for $350 million in Race To The Top funding for new assessments linked to common standards, reports Education Week. That means less multiple-choice testing  and more “essays, multidisciplinary projects, and other more nuanced measures of achievement.”

(The Education Department) wants tests that show not only what students have learned, but also how that achievement has grown over time and whether they are on track to do well in college. And all that, the regulations say, requires assessments that elicit “complex student demonstrations or applications” of what they’ve learned.

There is money for “comprehensive assessment systems” measuring mastery of a “common set of college- and career-ready” standards. Applicants get points for working with state universities to design the tests and guarantee that students who score above a certain level will be able to enroll in for-credit college classes.

Another pot of money will fund end-of-course high school exams.

Stanford Education Professor Linda Darling-Hammond, who leads a group representing a majority of states, believes performance assessments can improve the way teachers teach, notes John Fensterwald on Educated Guess.

The alternative is performance assessments, which require students to construct their own responses to questions. These can take the form of supplying short phrases or sentences to questions, writing essays or conducting complex and time-consuming activities, such as a lab experiment. “By tapping into students’ advanced thinking skills and abilities to explain their thinking, performance assessments yield a more complete picture of students’ strengths and weaknesses,” Darling-Hammond wrote.

“Performance assessments face obstacles of cost, reliability and testing time,” Fensterwald writes. He links to a critique of Darling-Hammond’s paper by Doug McRae, a retired publisher for the testing division of McGraw-Hill.

Because multiple-choice questions are cheap and easy to score, it’s possible to ask students a wide range of questions. As tests get more complex — write an essay, design an experiment, stage a debate — students  spend more time being assessed on far fewer prompts. Grading is subjective. Todd Farley’s Making the Grades explains tough it is for a group of people to score short answers and essays with consistency and fairness.

About Joanne


  1. If LDH is in favor of a test, she thinks it’s going to allow for subjective testing of low achievers that will close the gap. Count on it.

    Multiple choice questions do not simply test “low order” thinking. However, I would like to see multiple choice tests weight answers.

    For example, a common question presents a word problem that describes (and may or may not present) a situation that represents a right triangle and provides the length of the two legs. A basic question will ask “What is the distance from the coffee shop to the bakery?”–that is, a student has to calculate the hypotenuse.

    A more difficult question will ask, “If Joe goes directly from the the coffee shop to the bakery, how much shorter will his trip be than if he stopped at the laundromat first?” This requires the student to calculate the hypotenuse and then subtract that from the sum of the two legs.

    This question will ALWAYS include the length of the hypotenuse as one of the answer options. So a student who sketches the picture, realizes the method to solution, calculates the hypotenuse correctly, but fails to understand the import of the last question, is just as wrong as a student who adds the two legs together and selects that answer (also on the test) or who doesn’t know how to use the Pythagorean theorem.

    Rather than go LDH’s absurd route, I’d like to see us spend a bit more time weighting two answers. There’s often an answer that catches students who know the math but fail to read the question closely. That answer should get more credit.

  2. More bad news for bright but bored underachievers, students with language delays, non-native English speakers, students with organizational difficulties, and students on the autistic spectrum–including large numbers of gifted students for whom standardized multiple choice tests have often been the one remaining way to demonstrate academic skills.

    Katharine Beals

  3. Has anyone besides LDH done more to ruin the field of education in general and edschools in particular? Just asking.

  4. Katherine,

    Interesting comment and I’d like to know what your viewpoint is on the concept of ‘teaching to the test’ that so many states have adopted in order to deal with the amount of idiocy demanded by the federal government.

    As I recall, if a student learned the facts and concepts involved in a particular subject, they would be able to pass most any test given them (assuming they had truly mastered the material, as opposed to just memorizing answers).

  5. Bill,

    One problem with the tests as they are now, with respect to bright students, is that they set too low a bar and too low a ceiling. The result is that the brightest students aren’t able to distinguish themselves as they could back when the Iowa Tests were the standard. In my generation, you could be in 2nd grade and test at a 5th grade math or reading level. Now the best you can do is end up in the 99th percentile. A bright kid who makes stupid mistakes will have no way to redeem these mistakes, and demonstrate his/her potential, with correct answers on harder, above-grade level questions. The statistical “noise” created by stupid mistakes is a huge problem in a testing instrument that sets too low a ceiling.

    Then there’s the requirement on some (many?) of today’s state math tests that students explain their answers. Unexplained answers may only get half credit. Asked to explain answers to easy problems that he or she can do in his or her head, the bright under-achiever may not comply. Ditto if he or she has language difficulties.

    Teachers that teach to these kinds of tests will similarly lower the bar and the ceiling on classroom instruction, reducing the learning opportunities, the grades, and any other opportunities for academic distinction, of the brightest students in the class.

    Katharine Beals

  6. Katharine,

    I agree with your comments on the “explanation” issue but I disagree regarding the ceiling effect. There is, indeed, a ceiling effect, on most state accountability tests (incl. Calif.) but that is mostly as intended. Please remember that the annual state tests have generally no impact on individual students, and their main purpose is to assess the condition of the school with regard to “proficiency” on state standards. Consequently, that is the region (“interval”) where state tests tend to be most discriminating, as it is of the most importance to schools. Whether any given kid achieved at 98th percentile, or at 99th, or at 99.99th is of no real importance to anyone (except to the kid’s parents :-) What’s wrong with that?

  7. “Please remember that the annual state tests have generally no impact on individual students… Whether any given kid achieved at 98th percentile, or at 99th, or at 99.99th is of no real importance to anyone”

    Not so in Philadelphia, I’m sorry to report. Here, individual results are used to screen applicants to all the city’s selective middle schools and high schools. If you are below a certain percentile (85th or higher, depending on the school) you are automatically disqualified, no matter your grades, teacher recommendations, or other qualifications. Since Philly’s nonselective middle schools and high schools range from bad to terrible, the testing stakes are extremely high for us Philadelphia parents.

    I suspect that Philadelphia is not the only large school system to use low-bar state test results to screen applicants to selective high schools.

    Katharine Beals


  8. The ceiling effect does matter. If a substantial fraction of students are already able to work far above “grade level”, it means they are being under-served by the school and need more challenging material.

  9. I was one of those students who didn’t like to show how they got the answer on tests, but I had the ability to look at a problem and solve it (most times, correctly).

    Used to drive my teachers in math absolutely BONKERS. The problem with the school system is that many people forget that students (of all ages) learn differently (and if a 2nd grader is reading at a 4th grade level, doing math at a 6th grade level, and writing at at 3rd grade level, more power to the student).

    Testing has become the ‘method’ that schools want to use in order to determine how well their charges are doing, but there are students who don’t test well (which can be overcome, with proper training in test taking secrets).

    That being said, perhaps more testing isn’t the answer, but making sure students have mastered the basics of reading, writing, math, etc before leaving 5th grade.

  10. I’m actually a little shocked to learn that California’s state assessments don’t have any essay questions. Here in Colorado the tests have always had an essay component. LDH is not asking that multiple choice questions be eliminated, she just wants them combined with essay questions. The suggestion that tests should include some essay questions along with the multiple choice is hardly a radical idea. AP tests include essay questions without people making a huge fuss about it. It’s difficult to see how students could meaningfully demonstrate proficiency in writing if they’re not asked to write something.


  1. [...] This post was mentioned on Twitter by JoanneLeeJacobs. JoanneLeeJacobs said: New blog post: Race to new tests http://www.joannejacobs.com/2010/04/race-to-new-tests/ [...]

  2. [...] more: Race to new tests « Joanne Jacobs AKPC_IDS += [...]