Can Standardized Tests Save Education?

Posted on December 12, 2011

Photo by: Tilemahos_E

Scene: Coffee shop at convention center at 10:30 A.M., medications June 15, cost 2011.

Characters: Socrates and Dwayne Washburn, ailment Educational Policy and Management expert.

**** * Socrates: What do you think the most important current issue in education is?

Dwayne Washburn: I believe the Common Core Standards and the proposed Common Core Curriculum are very important, but I see the current problems with No Child Left Behind and schools not achieving adequate yearly progress as the most critical issues we face.

S: Would you say the problems are the result of a poorly conceived initiative?

DW: Absolutely not. The way AYP was formulated in the legislation for No Child Left Behind may have been ‘over ambitious’ but it served to put schools on notice that all students, regardless of their background, were to receive educational justice, which means that all should attain the same high standards.

S: As I understand it, all students are scheduled to pass the state test by 2014, regardless of what percentage of students pass the test now. Is that correct?

DW: Yes.

S: So, if a school has only 50 percent of its students passing the test now, it would have to improve twice as fast as a school that had 75 percent of its students passing. Is that correct?

DW: Yes. S: Isn’t this plan like a race between two runners with the faster runner closer to

the finish line than the slower runner?

DW: No. It’s not how much the students have to improve; it’s simply how many students in the school pass the test. The test is designed so it can be passed by any student who has an adequate grasp of the material for that grade. So all that is needed is instruction that brings more students to the passing criterion.

S: Are you saying all the students who don’t pass the test have the same number of incorrect responses?


DW: No, not necessarily. There are individual differences. S: And I presume that all who pass the test do not have the same number of

correct responses. Is that right? DW: Yes, of course.

S: So, wouldn’t it be possible to average the number of correct responses on a test to see how much total gain different schools would have to make for all students to pass the test? If the average deficit for one school was 3 items per student, wouldn’t that school be a lot closer to the goal than a school with an average deficit of 15 items per student?

DW: Yes, I suppose so. S: Doesn’t that difference mean that the school with the larger deficit requires a

lot more corrective instruction than the school with the smaller deficit? DW: Yes.

S: So do you still contend that it’s not how much students have to improve, simply how many students pass the test?

DW: I see your point. Very interesting. I don’t think I’ve ever heard of it expressed that way. (Makes note.)

S: And given that classes with larger deficits must go farther and faster to reach the finish line, where is the justice in this initiative?

DW: I don’t see that finding the average number who fail the test affects the moral mandate that at-risk populations must be taught more effectively.

S: I have no issue with the fact that students must be taught more effectively. I question the justice of AYP. Specifically, I question how much students are expected to learn in one school year.

Consider a 3rd-grade classroom that has only 40% of its students passing the third-grade math test. If only 40% are passing the test, wouldn’t the average student in this classroom have to learn some information and skills to pass the third-grade test?

DW: Yes.


S: And in addition, wouldn’t the average students have to learn a year?s worth of additional material to pass the fourth-grade test next-year?

DW: Well, I … I …

S: And if a third-grade classroom has 85 percent of the students passing the test, wouldn’t the average performer have less than year?s worth of material to pass the fourth-grade test?

DW: I’m not sure your idea of ‘the average student’ is valid. S: Let’s look at it another way, from the standpoint of teaching. Are the students

who are not passing the test the higher performers or the lower performers? DW: I think the answer is obvious.

S: Have these students ever learned more than a year’s worth of material in a school year before?

DW: Probably not. S: Yet, according to the mandate the teacher should be able to teach these

students more than a year?s worth of material in a year. Is this correct? DW: Yes, that’s what would have to happen. But …

S: No teacher in a typical at-risk school has ever taught the average student to learn more than a year?s worth of material in a year. Yet, this performance is expected by AYP. How is that possible?

DW: You make it sound as though the teacher is the only causal influence on the students’ performance, but in fact, the responsibility for the students’ performance does not lie entirely with the teacher. The influence of the home is of primary importance. The parents, the neighborhood, after-school programs—all these affect the outcome. And all should be involved in a coordinated effort to effect better learning outcomes.

S: Sweet words. But if the school fails to achieve AYP how much punishment is scheduled for the parents and the neighborhood? How much is history taken into account?

DW: Um … S: As I understand it, the school is the only agency punished.


DW: I wouldn’t call it punishment, but constructive reform. The school could be taken over, teachers or principals could be replaced with those who are able to meet the challenge of improving students’ performance.

S: And would those who serve as replacements be known quantities who have clearly demonstrated that they are capable of achieving more than a year’s growth in a school year? Or are they simply other people who may be no better than the ones they replace?

DW: You present a false dilemma. Replacements may not have a proven track record, but they could have the drive and motivation that makes them quite successful.

S: Is this reorganization plan one that has data? DW: Not that I know of, but it certainly should serve to clean up some of the

worst schools.

S: If you don’t have data, is your speculation about it happening a serious possibility or simply a wish?

DW: I resent that comparison. We know that if there are proper incentives, behaviors change considerably. With a good incentive program, some teachers would certainly be able to do it.

S: An interesting observation. Can you jump over a bar that?s six feet high? DW: Of course not. What’s your point?

S: I will give you one million dollars if you can do it. I’ll even give you three trials and let you prepare for one week. What do you think your chances are?

DW: Zero percent. S: Tell me why you are so skeptical that the attractive incentive wouldn’t make

the feat possible? DW: Because it’s beyond my capabilities.

S: So you’re saying that if something is beyond one’s capabilities, incentives can’t help them. Why isn’t that same rule applied to the teachers? They obviously don?t know how to achieve this goal.


DW: Individually, that’s true. But with a system that promotes teacher collaboration, whatever successful techniques are learned by one teacher would be shared with others.

S: Let me see if I understand this relationship. A teacher discovers something that is effective and then shares her discovery with the other teachers. How much time would it take for this teacher to teach the other teachers to do it properly?

DW: The time, I suppose, would vary with how complicated the technique is.

S: So it could take a long time, possibly hours.

DW: I suppose so.

S: Let’s say, the discovery relates to something in third-grade math. How many teachers would benefit from the discovery?

DW: Possibly one to three, depending on the number of third-grade classrooms.

S: So your plan is based on the assumption that 1 out of 4 teachers will make significant discoveries, and will be able to train the other three teachers on effectively using the discovered technique. Why don’t you use the same plan for students? Specifically, why don’t you just do away with the teacher, give students incentives for learning and have collaboration groups?

DW: I don’t see the parallel between teachers and students. S: Both are failures. You choose to teach students, but not teachers. Instead,

they are to discover. DW: Not so. The school and district provide in-service training.

S: But it obviously is not improving teachers’ performance. Here’s the basic problem, as I see it. If the students are not performing well, the most obvious remedy is to teach them using effective methods. If the teachers are not performing acceptably, teach them, using effective methods. Why don’t schools adopt that solution?

DW: Well, I think we’re getting far from AYP.

S: Then let’s look at things from the standpoint of a school that has fewer than 50% of the students passing the test. We’ll start with the core question that faces the school. Do you think this school would have to engage students in more “test prep” than it currently provides?


DW: Well, I would hope so.

S: Just so we understand each other: There’s a danger of going beyond what would be reasonable test prep. The test is valid only if students have learned a particular body of knowledge. The test simply samples items from that body of knowledge. If the test prep focuses on the specific items that will be tested, the test is no longer valid because it is no longer a sample; it becomes the body of knowledge that is taught. So would “legitimate test prep” focus exclusively on the skills that will most probably appear on the test?

DW: Of course, if it’s test prep it would focus on legitimate preparation for taking the test.

S: Does test prep include practice with items that are in the multiple-choice format, so that students mark a, b, c, or d?

DW: Yes. S: Would test prep have students practice with specific items that have been on

previous tests?

DW: I don’t know.

S: How much time would be devoted to test prep?

DW: I think some schools schedule it throughout the year.

S: And what is the relationship between the adopted textbooks and the test-prep work?

DW: I don’t know exactly what you mean?

S: The textbook was adopted because it was judged to be capable of teaching students the skills they need to learn. If a school devotes great amounts of time to test prep, it clearly will not have time available for teaching from the textbook. Do you see that as a problem?

DW: Yes, I see where it could be.

S: And here is the ultimate test-prep question: If you were a principal, would you condone or enforce test prep that provides students with practice items that are very similar to or identical to the items that will appear on the test they will take?


DW: You’re describing cheating. I certainly would not condone cheating in any way.

S: Consider the probable facts for your school. First, the probability is zero that your teachers in grades 2 through 5 could actually teach students what they would have to know to pass an unknown test that provides a fair sampling of items. The lower performers in the fifth grade would have to learn about three years? worth of material in one year to pass the test legitimately; the average performer, two years? worth of material

Next fact: Other schools that are similar to yours are getting high scores on this high-stake test. You know that their scores are bogus because students that transfer to your school from these schools don’t know any more than your students know. So other schools are definitely cheating. However, their officially recognized status is higher than yours because they have a higher percentage of students passing the test on all grade levels.

Next fact: You’ve been put on notice. If you don’t show at least 18% more of your students passing the test this year, you will lose your job and so will all the teachers. Most will probably be rehired, but you don’t want to lose your job and have to gamble.

Next fact: The district obviously is not trying to find cheaters because it’s in a position as hopeless as yours, with respect to possible takeover and reconstitution. So nobody is going to try to catch any over-ambitious test prep that you might provide.

Next fact: If your students learn the correct letter for all the items on the last four state tests, your students will probably pass the test. This goal is possibly within reach, at least for many of your students.

Final fact: If you spend all year with test prep as your primary focus, you won’t really teach your students much about math, but fourth-grade students who perform on a second-grade level won’t learn much anyhow.

Given these facts, tell me you would not close your eyes to cheating or even engage in some subtle promotion of illicit preparation.

DW: If the information you present as facts is true, I would be tempted to cheat, but do you really think that many schools engage in this kind of cheating?

S: I’m convinced of it. But there’s a simple way to prove me wrong. What is that way?


DW: (Pause). Possibly give students a second test and see if they perform as well on that test?

S: Exactly, ideally a normed test that covers the same scope. If a large percentage of students pass the state test but fall below the 50th percentile on the normed test, cheating is probably occurring.

DW: I find your evaluation hard to accept. S: But is the problem with the teachers and administrators or with the

discriminatory nature of AYP? DW: How is AYP discriminatory?

S: Here’s a question: Why don’t we give the fifth-grade test to all students who are in third grade?

DW: Probably because it’s a fifth grade test. S: Does that mean you think it wouldn?t be fair for students who perform at the

third-grade level?

DW: Right. You can’t expect children who are in the third grade to pass the fifth- grade test.

S: Can you expect fifth-graders who perform at the third-grade level to pass the fifth-grade test?

DW: Probably not.

S: And therein lies the discrimination. If the average student in a fifth-grade classroom performs at the third-grade level, it is as discriminatory to give students the fifth-grade test as it is to give the fifth-grade test to younger children who perform at the third-grade level.

DW: I may be missing the point, but I don’t see much discrimination in giving a fifth-grade test to fifth graders.

S: It’s not simply the test; it is the implications of failing the test. Consider math. Fifth-grade students who perform on the third-grade math level take the fifth- grade test this year. Which test will they take next year?

DW: The sixth-grade test.


S: And is the preparation they receive next year based on third-grade math or sixth-grade math?

DW: Sixth-grade. S: And how much of this do you think would be comprehensible to the students? DW: Probably not everything.

S: Here’s a sixth-grade math problem: If the factory produces 14 containers every 5 seconds, how much time would it take the factory to produce 322 containers?

To solve this problem, a student would have to understand ratios, would have to work the two-digit division problem, 322 divided by 14, and would have to multiply the answer by 5. Isn’t it discriminatory to require students who do not have these skills to have to work such problems?

DW: Yes, if you put it that way. But you seem to have a way of making the whole thing look pretty ugly.

S: But if it is actually ugly how can you address it with realistic remedies if you don’t acknowledge the ugliness?

DW: But if the whole scheme is as ugly as you portray it, how can we possibly fix it?

S: There are two choices: affirmative action or placing students in the grade level that corresponds to their skills and knowledge.

DW: I don’t think it would work well to place fifth graders in the third grade.

S: Then there has to be some form of affirmative action.

DW: What would that be?

S: Is it fair to say that we should gear the instruction to the skill level of the students and not try to teach things that are far above their skill level?

DW: Yes.

S: So we would place students in instruction based solely on their skill level. Next, we would use programs that have been demonstrated to teach average at- risk students more than a year?s worth of skills in a year.


DW: So you would require students to learn at an above average rate?

S: Yes, that’s part of the affirmative action. But obviously this can’t happen with any program. It has to be one that has been shown to be effective. Also, it places students where they belong in the program sequence, which means that they would know possibly more than 80% of what’s covered in each lesson, and would have to learn less than 20% new material on any lesson. Note that the primary evaluation would not be on the state test. It would consist of weekly or bi-weekly tests of students’ performance in the instructional program. It would show how all students perform on everything that is being taught.

DW: Why would these be the primary source of data?

S: Because students are to learn everything that is taught not simply a few multiple-choice items. The tests would have no multiple-choice items but, rather, items that provide good evidence of student performance in the program. With timely data we can respond constructively to problems. If we don’t have the tests and other classroom observations, we don’t have the information that is necessary to provide timely responses to problems.

DW: So you would provide this structure only for failed schools?

S: If it were my choice, I would provide it for all schools. But given the unlikelihood of that possibility, I would put the effort where there is the greatest call for justice—with our at-risk students.

DW: Do you think the at-risk population would catch up to the regular population as they move through the grades?

S: If the regular population was taught very poorly, and the at-risk population very well, the at-risk population could easily pass up the regular students. But wouldn’t that be a system as discriminatory as the one that currently exists?

DW: Yes, I believe it would be. But if the regular population is taught better than it is now, the at-risk population would never catch up.

S: Here’s a question: If a person had the skills necessary to become an engineer, and if this person had a strong work ethic and good ability to learn new things, would it matter that some other persons have more skills than this person?

DW: I don’t think so. It’s not just the skills you have but how you use them.


S: So the issue is not whether the at-risk population catches up. The issue is whether this population has the skill set necessary for the full range of future schooling and employment. In other words, if the schools bring both populations to the highest standards that are in reach, the at-risk population would perform higher than today’s regular population. The more advantaged population would be ahead of them, but we would clearly have a system that metered out justice for all.

DW: (Makes note and stands up.) I have a lecture to give in 15 minutes. Thank you for sharing your insights with me.

Reprinted with permission

Author: Siegfried Engelmann

Photo by: Tilemahos_E  (Flickr User)

Read more from Siegfried Englemann HERE>

The myths of standardized testing

Are the following statements true or false?

  •  Students’ knowledge and skills can be assessed by a sample of content that makes up a 45-question test.
  •  High test scores of students at any particular school prove that there is high student achievement and quality teaching at the institution.
  •  Punishments or rewards to teachers or students based on test scores motivate them to do better.
  •  A standardized test score is a better reflection of student learning any any other form of assessment.
  •  If the stakes to a test are high enough, viagra buy people will work harder and improve their performance to meet the challenge.

These are common myths of high-stakes standardized tests, clinic which have become the focus of modern school reform and used to evaluate schools, students and, increasingly, teachers. Research…

Author: Valerie Strauss, The Washington Post

Read more HERE>