Education Trends Subscribe to RSS

A Different Perspective: Teaching to the Test

| Ben Johnson

What does "teaching to the test" mean? I haven't actually ever seen this literally happen in a classroom: "Class, remember the answer to question 12 is A, 13 is B, and 14 is D." But, as much as it is maligned, isn't a form of teaching to the test the point of why we teach in the first place? I'm wondering how students can be successful on the state standardized test if we don't teach to it?

An Analogy

Gideon, my son, plays on his high school soccer team. He plays one of the two fullback positions. Because of his large size and speed, he is able to defend the goal from the other team successfully, primarily because he knows where the goal is and he knows the purpose of the ball. The test for him is to keep the ball out of his own goal and get it into the goal of the other team. He has to know the rules about not going further than the last defender of the other team or he will be off sides. He has to know that he cannot touch the ball with his hands, or kick it dangerously. All of this would be useless for Gideon if he did not have the stamina to run, or the skills to control and kick the ball.

Not teaching to the test is similar to the coaches constantly drilling the players on dribbling the ball, kicking the ball, and being able to run fast, but none of the players ever being told the purpose of the goal or why it would be important to defend it. Can you imagine the total hilarity of such a game?

Imagine twenty-two players frantically kicking a ball, running, and then kicking some more, not concerned about what direction the ball is kicked. Some players will enthusiastically run and follow the ball, while others will stand idly by and wait for the ball to come to them. Some players might even just give up in frustration and lie down on the field. If the ball accidentally makes it into the goal, the crowd cheers, but the players don't have a clue as to why.

Building Stamina and Providing Purpose

Does this sound like any classrooms we know about? Are we guilty of being so concerned about not teaching to the state standardized test that we make the students spend most of their time answering multiple-choice questions just to get them ready? What about the hours and hours we spend on showing the students how to eliminate wrong answers in order to increase their chances of getting the right ones?

Does it bother us that many students lie down on the field in frustration because the test doesn't mean anything to them? How long will students enthusiastically run after a bubble sheet if they do not know what to do with it and if they don't know why they are taking it? My point is, students should know exactly what is on the test and exactly why they need to know those things, otherwise, how in the world do we expect them to be successful on a test?

Why would we teach to something other than the test? If we are not teaching to the test, to what are we teaching? Education leader and researcher Fenwick English emphatically states that there is no shame in teaching to the test as long as the test is rigorous and representative of what needs to be learned. Such a curriculum-based test is prepared in advance as Grant Wiggins and Jay McTighe suggest in their backwards-planning guide, Understanding by Design: The teacher and the students both know exactly what the test is and they work together to meet the standards of the test.

Tests are facts of life, we deal with them because we know that somewhere, someone has to draw the line and set a standard. What are your thoughts about teaching to the test?

see more see less

Comments (69)

Comment RSS
Education Consultant dedicated to improving schools, one teacher at a time

Shell game of testing

Was this helpful?
+1

James Mason is right on target. The title of the blog is misleading. The knee-jerk reaction is that we should not teach to the test. I simply followed that line of thinking, especially from Fenwick English (2000) who says that keeping any part of the curriculum out of the instruction skews a test, so the goal is to get every part of the curriculum evenly to every student in a manner that they can be successful at acquiring that set of knowledge and skills. Then the bell curve becomes meaningless because it is not a curve at all, it is more like a"J" (p. 74). Simply put by English, the written curriculum should match the taught curriculum which should match the tested curriculum. The only way that can truly happen is with the local curriculum, local instruction and local tests--i.e. what the teacher does in the classroom. That is where the focus should be, not on the state standardized minimum performance test.

Great post-- Thanks for the research!

Reference
English, F. (2000). Deciding what to teach and test. Thousand Oaks, CA: Sage Press, Inc.

Ben Johnson
San Antonio, TX

Quote:

To address the charge of what actually happens in schools, teaching "12 is A, 13 is B" would not be teaching to the test, it would be cheating. And while teacher cheating DOES occur in schools (Amrein-Beardsley, Berliner, & Rideau, 2010), it is not as common as we would think based on the scandals that are currently erurpting in the media.

What is commonly called "teaching to the test" neither that kind of blatant cheating, nor is is what Ben Johnson describes… which is not teaching to the test at all but teaching to the underlying goal of education (which is NOT the test). The literature describes three different distortions of the curriculum relevant to this discussion:

Narrowing of the Curriculum:This consists in focusing resources (mostly instructional time) on the tested subjects (such as Reading and Math) to the detriment of non-tested subjects. This is what many people mean by "teaching to the test", and it is a serious concern to those who value a broad education. It is also well-documented in the literature, with Au (2007) being a good place to start.Teaching to the Test:All areas of a tested subject do not carry equal weight on the test. Since there are many more topics taught in one year of mathematics than can fit in a 72-item multiple choice test, test developers are forced to take a _representative sample_ of the tested domain when developing their tests. (See Koretz, 2008 for more detail on the sampling issue.) Once this sample is known, teachers and administrators can take advantage of this by focusing instruction on only the topics known to be tested. This destroys the representativeness of the sample, and means that student performance on the test can no longer be extrapolated to knowledge of the broader subject area. Teaching this way tends to reduce each subject area to isolated facts, and has been documented as far back as Smith's (1991) classic study of test preparation.Teaching the Test Itself:This practice consists in using real or simulated test items ASinstructional materials. The overuse of Released Test Questions, as well as the test-prep packages sometimes mandated by administration, falls into this category. In this form of test prep, the test as become "the final purpose, the ultimate goal" of education… something the New York State Department of Education warned us about in 1906."Why would we teach to something other than the test?" Because even the best test is just a small sample of what we want students to know, and also just a small sample of the content standards that we are required to teach. Teaching to the test robs students of the "purpose" that Ben Johnson rightly advocates, since the purpose of education is NOTthe test.

Just a few references on the large topic of test prep:
Amrein-Beardsley, A., Berliner, D. C., & Rideau, S. (2010). Cheating in the first, second, and third degree: Educators' responses to high-stakes testing. Education Policy Analysis Archives, 18(14).
Au, W. (2007). High-Stakes Testing and Curricular Control: A Qualitative Metasynthesis. Educational Researcher, 36(5), 258–267. doi:10.3102/0013189X07306523
Koretz, D. M. (2008). Measuring Up: What Educational Testing Really Tells Us.Cambridge, MA: Harvard University Press.
Smith, M. L. (1991). Meanings of Test Preparation. American Educational Research Journal, 28(3), 521–542. doi:10.3102/00028312028003521

Physics & Engineering Professor

I agree - comparing tests to soccer is a poor metaphor

Was this helpful?
+1

I couldn't agree more with Mitch. A test cannot ascertain is a person is a good soccer player any more than standardized tests ascertain is a student has mastered his or her subject matter.

Education Consultant dedicated to improving schools, one teacher at a time

Soccer and Testing

Was this helpful?
+1

Mitch:

You make a point about "soccer" not being a good analogy. What would be a better one to show the futility of learning and learning, practicing and practicing without knowing why? I thought the point was well stated... Tests, whether standardized or curriculum based, have to mean something to motivate the best performance from the students as they prepare to show what they know.

Yes, there are inherent problems with standardized tests that attempt to sample student learning progress. The real power of testing has to be harnessed by the teachers in the classrooms. Frankly, if teachers were doing a good job with gathering descriptive evidence of student learning in the classroom (I am not talking grades, I am talking about showing student performance on established learning objectives), then there would be no need for the national movement for state standardized tests.

Ben Johnson
San Antonio, TX

Education Consultant dedicated to improving schools, one teacher at a time

Only the test

Was this helpful?
+1

AShaw:

Makes a very good point. Standardized tests are minimum tests and teaching to that tests is eliminating the full curriculum.

Another excellent point made is that it is not the test, but the teachers that are reacting to the test that are causing the problem. When stressed, teachers fall back on what they are comfortable with. The most comfortable (easy) method is worksheets and drill and kill. What suffers are the strategies that we know increase student academic performance and student motivation. How do you motivate a student to do well on a worksheet?

Ben Johnson
San Antonio, TX

Quote:

Teaching to the test does not help produce productive citizens and life long learners. Many students (including English Language Learners- ELLs and students with learning disabilities) have difficulty with the lingo of a test and are simply not 'good' test takers. Tests do not take into account our students background knowledge or schema. Test do not take into account our students interests either. Howard Gardener's multiple intelligence theory shows that we have different learning styles and strengths. Summative assessments are not the only measure of our intelligence. The idea of starting 'practice' standardized tests every day for 2-3 months before state assessments makes me sick to my stomach. Many teachers are drilling and strictly focusing on the standardized assessment. This is disregarding the overall goal of developing well-rounded citizens. Think about why you became a teacher. Did you want to make a difference in children's lives or the difference in 'meeting the standards?'

Yes, we do need to teach the content and it would be beneficial to teach test taking skills. Should we make make 'teaching to the test' our focus? Teaching your students the state standards and teaching your students practice tests is a totally different concept.

You cannot teach what is ONLY on the test. When a child comes from another country, you not only have to teach content but also the English language. When a child comes from a rough home life, you cannot teach what is just on the test because the child needs so much more than content. A teacher is not there to teach just standards or content, but teachers are there to ensure that their students are well rounded in every aspect of life.

"No Child Left Behind's emphasis on 'teaching to the test' undermines quality teaching" http://www.newsdesk.umd.edu/culture/release.cfm?ArticleID=1576

Educating the whole child is MORE than 'teaching to the test'

Was this helpful?
+1

Teaching to the test does not help produce productive citizens and life long learners. Many students (including English Language Learners- ELLs and students with learning disabilities) have difficulty with the lingo of a test and are simply not 'good' test takers. Tests do not take into account our students background knowledge or schema. Test do not take into account our students interests either. Howard Gardener's multiple intelligence theory shows that we have different learning styles and strengths. Summative assessments are not the only measure of our intelligence. The idea of starting 'practice' standardized tests every day for 2-3 months before state assessments makes me sick to my stomach. Many teachers are drilling and strictly focusing on the standardized assessment. This is disregarding the overall goal of developing well-rounded citizens. Think about why you became a teacher. Did you want to make a difference in children's lives or the difference in 'meeting the standards?'

Yes, we do need to teach the content and it would be beneficial to teach test taking skills. Should we make make 'teaching to the test' our focus? Teaching your students the state standards and teaching your students practice tests is a totally different concept.

You cannot teach what is ONLY on the test. When a child comes from another country, you not only have to teach content but also the English language. When a child comes from a rough home life, you cannot teach what is just on the test because the child needs so much more than content. A teacher is not there to teach just standards or content, but teachers are there to ensure that their students are well rounded in every aspect of life.

"No Child Left Behind's emphasis on 'teaching to the test' undermines quality teaching" http://www.newsdesk.umd.edu/culture/release.cfm?ArticleID=1576

Comparing learning how to

Was this helpful?
+2

Comparing learning how to play soccer to learning how to take a standardized test does a huge disservice to the game of soccer. Begin with motor skill development. One can learn every rule and strategy for soccer, but without well developed motor skills that person will still be a terrible soccer player. This same person might easily ace a standardized test about playing soccer without ever stepping on the field.

Along with the need for excellent motor skills, a good soccer player needs excellent problem solving skills that reflect the highest levels of Bloom’s Taxonomy. Every game is filled with novel situations which include an infinite number of variables. Relying solely on memorization and recall will have dismal results on the soccer field. Memorization and recall are of preeminent importance when it comes to doing well on a standardized test. These skills fall on the lowest rung on Bloom’s Taxonomy.

There is nothing analogous about playing soccer and taking a standardized test, just as there is nothing analogous about taking a standardized test and solving real world problems.

Citizen who understands Education is the most important piece in society.

You assume the tests are valid

Was this helpful?
0

You are begging the question. You are assuming teachers are teaching to tests that are valid. They are not valid tests and THAT is why teaching to tests is a problem. Among other issues, standardized tests do not test how a student innovates with her collective knowledge and creativity.

Teaching to the standardized test is the equivalent to the coach saying, "Trust me that passing the ball is an important part of soccer. Regardless, the State requires you know how to pass the soccer ball a certain way, so we are just going to work on passing all this week. As such, we don't have time to play the game. We don't have time to incorporate all the different skills you learned together because you will not be tested on that. You will only be tested on each skill, independent from one another. Each week, we will focus on a new skill (and never play the game)." If you have not seen THAT happen in a classroom, visit more classrooms.

You make it sound simple by implying that the test should be scoring, but that is not good either. Scoring could just mean the other team's goalie is bad or we have one really good player on our team. How does the score reveal how good Gideon the fullback did on passing, heading the ball, playing zone defense, and not using his hands, particularly when he was just one of many players? If scoring was the best indicator, college recruiters would just wait to see what team wins at the end of the year and pick all the kids from that team first. The recruiters know it is not that simple.

How do you test how well an individual player did? That is a tough question and it is one that the standardized state tests do not address. They might be able to test how well he passes, kicks, or heads the ball just like a student might be good in English, Math, and History. But it doesn't say how good the player is just like we don't test how well a student innovates using English, Math, and History together. In particular, a good soccer player uses finesse and creativity, leverages opportunities, and works as part of a team - all of which are hard to measure and are not tested in the standardized tests that I know.

Quote: What about the driving

Was this helpful?
0
Quote:

What about the driving license? Isn't a test.A test, in my opinion, could test discrete points/items,or a construct( comprehensive); this depends on the purpose of the test.

Bringing the drivers license into a debate about standardized testing is usually a red herring. Never the less, could you imagine licensing drivers based on ONLY the written test and not the road test?

Back to basic testing theory: an objective test a cannot measure any large construct directly, because it relies on discrete items it necessarily will be testing discrete skills. Because of this, such a test cannot assess a large domain exhaustively, without sampling.

If you wished to assess a child's knowledge of the alphabet, it would be feasible to test the entire domain of 26 letters. However, if the domain were "fourth grade mathematics", the number of items needed on the test would be unmanageable because of the breadth and depth of the material. In order to make the test of reasonable length, some tradeoffs have to be made, and not all the content would be covered by the sample. That would be fine without the high stakes; but once the pressure is on, it is inevitable that people will "game the system" by teaching to the test.

Founder and Director at Jude for Learning Enhancement

What about the driving

Was this helpful?
+1

What about the driving license? Isn't a test.A test, in my opinion, could test discrete points/items,or a construct( comprehensive); this depends on the purpose of the test.

Distortion of the Curriculum

Was this helpful?
+1

To address the charge of what actually happens in schools, teaching "12 is A, 13 is B" would not be teaching to the test, it would be cheating. And while teacher cheating DOES occur in schools (Amrein-Beardsley, Berliner, & Rideau, 2010), it is not as common as we would think based on the scandals that are currently erurpting in the media.

What is commonly called "teaching to the test" neither that kind of blatant cheating, nor is is what Ben Johnson describes… which is not teaching to the test at all but teaching to the underlying goal of education (which is NOT the test). The literature describes three different distortions of the curriculum relevant to this discussion:

Narrowing of the Curriculum:
This consists in focusing resources (mostly instructional time) on the tested subjects (such as Reading and Math) to the detriment of non-tested subjects. This is what many people mean by "teaching to the test", and it is a serious concern to those who value a broad education. It is also well-documented in the literature, with Au (2007) being a good place to start.
Teaching to the Test:
All areas of a tested subject do not carry equal weight on the test. Since there are many more topics taught in one year of mathematics than can fit in a 72-item multiple choice test, test developers are forced to take a _representative sample_ of the tested domain when developing their tests. (See Koretz, 2008 for more detail on the sampling issue.) Once this sample is known, teachers and administrators can take advantage of this by focusing instruction on only the topics known to be tested. This destroys the representativeness of the sample, and means that student performance on the test can no longer be extrapolated to knowledge of the broader subject area. Teaching this way tends to reduce each subject area to isolated facts, and has been documented as far back as Smith's (1991) classic study of test preparation.
Teaching the Test Itself:
This practice consists in using real or simulated test items AS instructional materials. The overuse of Released Test Questions, as well as the test-prep packages sometimes mandated by administration, falls into this category. In this form of test prep, the test as become "the final purpose, the ultimate goal" of education… something the New York State Department of Education warned us about in 1906.

"Why would we teach to something other than the test?" Because even the best test is just a small sample of what we want students to know, and also just a small sample of the content standards that we are required to teach. Teaching to the test robs students of the "purpose" that Ben Johnson rightly advocates, since the purpose of education is NOT the test.

Just a few references on the large topic of test prep:

  • Amrein-Beardsley, A., Berliner, D. C., & Rideau, S. (2010). Cheating in the first, second, and third degree: Educators' responses to high-stakes testing. Education Policy Analysis Archives, 18(14).
  • Au, W. (2007). High-Stakes Testing and Curricular Control: A Qualitative Metasynthesis. Educational Researcher, 36(5), 258–267. doi:10.3102/0013189X07306523
  • Koretz, D. M. (2008). Measuring Up: What Educational Testing Really Tells Us. Cambridge, MA: Harvard University Press.
  • Smith, M. L. (1991). Meanings of Test Preparation. American Educational Research Journal, 28(3), 521–542. doi:10.3102/00028312028003521
see more see less