What’s Really Wrong With Standardized Tests

I've always been agnostic about all the state testing requirements that came out of NCLB. I didn't think that they should be used to evaluate teachers' performance, but I thought that they were very useful for seeing comparing performances between richer and poorer school districts.

Statistics are always best when they are measuring a large sample of individuals. A greater N, the more reliable your findings. That's why evaluating a teacher using standardized tests is useless. Because N = 25. A teacher shouldn't be evaluated based on such a small sample of kids. 

I didn't even mind that kids lost four days to take the test. It seemed harmless enough, when it could provide such useful data. I didn't mind that some districts were teaching to the test, because those districts could use a little guidance any way. 

However, over the past few years, it has become increasingly obvious that these tests are being used in a really nasty way, and now, I'm majorly rethinking everything that I've ever said about standardized tests. 

While districts are fighting every effort to use standardized test to measure teachers, they have, at the same time, been using these tests to evaluate students. These tests, which are supposed to give us a slice of information about a large group of kids, are being used to place individual kids in honors classes and in special education classes. 

My son's classmate received As and Bs in his English class, but when he scored on the low average range of the standardized test, he was placed in a basic skills English class. The principal ignored an entire year's worth of classwork and instead used that one test to determine his placement. I've also seen the test used to determine higher placement, as well. It determines placement in honors class and the Gifted and Talented program. 

So, now these tests take precedence over schoolwork, even though the tests were designed to only measure a very minor aspect of school learning, on one particular week when millions of variables could interfere with the results. 

Because schools are under such pressure to raise their test scores, which determine rankings and even property value, there is growing attention on those that finish in the top 10% percentile. It's this group of highly intelligent, highly motivated, highly prepared students that help to determine the overall average for the school. One high scoring kid can kick that average up. (And the 15% of special ed kids can really screw with the average.) 

So that group of high fliers are pampered by the school district. They are fought for by competing charter schools and private programs. During the high school orientation last week, the principal explained that we were losing about 8% of the middle school population to magnet and private schools and they were going to start recruiting kids to stay in the school system when they were in 7th grade. Losing 8% of the town's smartest kids was a huge issue. Because it fucked with their averages. 

It bothers me that a school district places all of its energy catering to one sample of the population. They should be working to provide great teachers to the average student, too. Why doesn't every kid get the cool lessons and brain teasers in the Gifted and Talented program? 

I've been writing all weekend about the stress in the teenage world. The mis-use of standardized testing is part of the problem. 

35 thoughts on “What’s Really Wrong With Standardized Tests

  1. Anecdata ahoy: My kids are about to start a stock market unit in their fifth grade (private school) classes as part of their math studies–I did a similar project in fifth grade (public school), but mine was only for a small group of G&T fifth graders, while their whole grade is participating. They also have about 15 students in their homerooms, uncertified-and-yet-still-excellent teachers (for the most part, private school teachers are not certified) and no standardized testing.

    Like

  2. To combat that, Michigan also looks for the “achievement gap” between low and high scorers on the test. Too much deviation is also seen as “bad” with many of the more affluent districts put on a watch list that a more uniformly low scoring district would not.
    My son qualified for special education services by extensive psychological testing, classroom observation, and behavior charting; however, he is a very high achiever on the standardized assessment so it was determined that he would not receive services.
    The educational system is not rational and will not produce a quality citizenry under the various NCLB mandates and privatization schemes that have been implemented. I see very little room for individual creativity – either student or teacher initiated under the current compliance heavy model.

    Like

  3. “My son qualified for special education services by extensive psychological testing, classroom observation, and behavior charting; however, he is a very high achiever on the standardized assessment so it was determined that he would not receive services.”
    This is the ace I keep up my sleeve. If I get pissed at the school district, I may opt my son out of testing. Buh-bye perfect math score, bringing the averages up!

    Like

  4. As a middle scool principal, I can tell you that the principal you describe has got it backwards. If a school focuses on kids learning what they are supposed to learn and supporting those who for whatever reason struggle, the test scores will go up. The best way to get test scores to go up, is to ignore the test and focus on proven ways to help kids learn. And if you are placing students in a remedial class without at least five pieces of reliable data to back it up, including teacher recommendation, you are asking for trouble.

    Like

  5. I think this Washington Post reprint of a letter from a high school teacher to college professors identifies the problems with our current educational testing culture very effectively:
    http://www.washingtonpost.com/blogs/answer-sheet/wp/2013/02/09/a-warning-to-college-profs-from-a-high-school-teacher/
    His observation that “the drivers of the policies that are changing our schools—and thus increasingly presenting you with students ever less prepared for postsecondary academic work—are the wealthy corporations that profit from the policies they help define and the think tanks and activist organizations that have learned how to manipulate the levers of power, often to their own financial or ideological advantage” is especially relevant.

    Like

  6. In my school district, standardized test scores are one of a several factors considered for advanced, accelerated, or gifted learning. They also consider the teacher’s recommendation, grades in the class, the child’s motivation, etc. That way, if a kid doesn’t test well, but the teacher knows he or she can do the work, they can still be promoted.
    Not that I’m defending standardized tests in any way… I hate the damn things. But standardized tests aren’t evil if a school district has some perspective about them. An A or B student who stinks at testing shouldn’t be penalized for a poor grade on a standardized test. The school should use a little common sense.

    Like

  7. I like tests because I like measuring things (if it doesn’t take too long to measure them). I do agree that tests are often used to measure badly — measuring someone’s height with a yardstick, believing you know their weight, or using their shadow to measure their height without knowing what time of day it is. And, even when measured well, they are put to bad uses.
    I also further admit that if the bad uses become bad enough, that one has to give up the measurement (for example, I wouldn’t mind knowing the heights of kids in a class, but if that’s used to assign them to math class, it would be a bad enough thing that we have to advocate no longer measuring their heights).
    On the other hand, I do think that NCLB exposed an underclass of education that was only pretending to educate. This exposure hasn’t produced education, but at least it forced us to confront the situations in which we actually were doing poorly.
    I think the answer is low stakes testing, but even in that circumstance, it’s hard to avoid people attaching more importance than they should to the tests.

    Like

  8. I do not thing “stress” is sufficient reason to stop testing. To use terms from the Bronson NY Times article, some amount of stress (i.e. “challenge” rather than “threat” — he likes making up words for categories) is a good thing.
    Can you give some examples of how the top % is being pampered in your school?
    Do they get smaller class sizes, more field trips, other treats that everyone would want?

    Like

  9. bj: ” I do think that NCLB exposed an underclass of education that was only pretending to educate. ”
    Yes, I totally agree. That’s why I liked (note the past tense) NCLB and the standardized tests. But that’s not how those tests are being used by individual school districts.
    In the past two years, we’ve been in two separate school districts – one high achieving and one that was less achieving. Both of those school districts use the state standardized tests in the WRONG way. When they are used to give preference or penalties to individual kids that is the wrong way.
    Like I said, a large sample size = reliable findings. In the case of these tests, we’re looking at one kid, one day, one test. N = 1. Principals are using those tests and not an entire year worth of information from teachers based on tests, homework, and class participation to make huge decisions about tracking.
    And, so you all know that this isn’t sour grapes, let me tell you one story about Jonah. Going into 6th grade, Jonah was going to be placed in a class for High-Average math kids. He had been getting B+’s in his math class in 5th grade, because he would forget homework or do silly mistakes on his tests. He would forget to label his answers or do all the math in his head and not show the work. Whatever.
    I went in to see the principal to find out how he did on the state standardized test. When she looked up his score and found that he got a perfect test score, he was immediately placed in a high math class. She did not consult the teacher or anything.
    In this case, my kid benefitted from a standardized test, which didn’t care if you put down a label or showed your work. Of course, that makes me happy, but the point is that these tests matter. A lot. Too much.
    Another anecdote. I went to a PTA meeting in the new school district about a year ago. The principal talked to the parents about the results of the latest test. For some reason, the kids in the school district didn’t do well for the 6th grade test, but jumped up in the 7th grade. After a couple of seconds of speculating about why this was the case, he told the room of moms that that it didn’t matter. The tests shouldn’t tell us anything about the school district, they only told you something about the kid. Which was totally and incredibly false.

    Like

  10. This emphasis on testing is destroying education. If parents don’t wake up and get organized, students will continue to suffer the consequences. Politicians and the corporations are making a tremendous amount of money. Teachers are being publicly ridiculed and their opinions are being ignored. Parents must stop this!

    Like

  11. Well! Welcome to the other side!
    I want to edit your sentence: “Losing 8% of the town’s SMARTEST kids was a huge issue.” to:
    “Losing 8% of the town’s HIGH-STANDARDIZED-TEST-SCORING kids was a huge issue.”
    Anyway, I’m assuming that you’ve read Alfie Kohn’s anti-testing, anti-competitive books and articles and I’m wondering what you thought about them. I’d really like to read a review of some of his books by you, to see what you think/feel about his approach to children’s education.

    Like

  12. “I went in to see the principal to find out how he did on the state standardized test. When she looked up his score and found that he got a perfect test score, he was immediately placed in a high math class. She did not consult the teacher or anything.”
    That’s a very common story from the old days, I think–a diamond in the rough who only gets discovered by his school because of his standardized tests.
    “The principal talked to the parents about the results of the latest test. For some reason, the kids in the school district didn’t do well for the 6th grade test, but jumped up in the 7th grade. After a couple of seconds of speculating about why this was the case, he told the room of moms that that it didn’t matter. The tests shouldn’t tell us anything about the school district, they only told you something about the kid. Which was totally and incredibly false.”
    That sounds like there was a problem in 6th grade.
    Actually, come to think of it, maybe the problem with 6th grade was that the kids were losing ground just because of the shift from elementary to middle school. Maybe by 7th grade, the kids had successfully acclimated.

    Like

  13. Miranda, two of my kids have had the opposite happen. They both tested really high in standardized tests and IQ tests, but there daily work was lagging. They qualified for services and IEPS to help bridge the gap. It was totally related to maturity for my daughter, she moved on and became a really good student all around. She is a freshman in college now. My son still struggles with attention and coordination problems. Still qualifies for services, but he can ace their tests.

    Like

  14. I am no enthusiast for American secondary education; my daughter’s very expensive (Manhattan private school) education produced fairly mediocre results. But I’m not sure what wealthy corporations have to do with it. Also, I think that high school courses in government are part of the problem. High school students should spend full time on traditional liberal arts subjects. Above all, they should learn to write clearly and effectively, which requires weekly graded essays, returned with lots of red ink reflecting lots of attention to prescriptivist rules.

    Like

  15. “Above all, they should learn to write clearly and effectively, which requires weekly graded essays, returned with lots of red ink reflecting lots of attention to prescriptivist rules.”
    We have a college-wide writing assessment, examples of which I have been reading all weekend, and it still amazes me how students write so incorrectly. I am starting to wonder if in K-12 we should be focusing solely on grammar and organization and not on critical thinking, then letting us in college work on critical thinking. I sometimes feel like K-12 and college are trying to do both and doing half-assed jobs at both. (Or rather, in college we end up having to do both because K-12 is doing a half-assed job at both.)

    Like

  16. “That’s a very common story from the old days, I think–a diamond in the rough who only gets discovered by his school because of his standardized tests.”
    We’re hearing this is still a story, though, and, it’s one of the benefits of standardized testing: Kids who aren’t as good at some of the other demands of school (i.e. sitting still, interacting with the teacher, doing their homework neatly and on time, keeping track of due dates) or whose teachers are biased against them (consciously or unconsciously) can get access to their educational level.
    That said, I am also comfortable with the notion that a a tool I might find useful is in practice used so poorly and badly that we can’t use the tool (even though I might support it in theory). To show me that, I’d need to see evidence, that the costs outweighed the benefits to evolve my views on testing. I’d also need more than just a generalized fear of stress (though there, too, I’m willing to look at evidence). Otherwise I find the emphasis on stress avoidance/self esteem/how you feel as the dark flip side of the dark side of an over-reliance on testing, a flip-flopping backlash, and not the middle ground I’d like to find, where tests are used wisely. And, it’s not just tests, but the idea that there is content to learn and teach and that we have to have ways of assessing that learning.
    By “evidence” I think I mean talking about our goals for education and then whether our practices are reaching those goals or not, which is by no means a simple problem and not just a research problem (since setting goals is not answered simply by the data but by the values we attach to different endeavors).

    Like

  17. “High school students should spend full time on traditional liberal arts subjects.”
    I’d probably say a lot more emphasis on math (including the branches of math that have become more important in modern society, like probability and statistics and data analysis and visualization). Then I’d add in science with an emphasis on querying the world scientifically (which doesn’t mean without content), standard usage of English, then I’d add something that teaches kids to synthesize information and explain about it.
    My impression is that synthesis and explanation is really what many are talking about when they say “writing” (along with standard usage). Going into the future, I’m not sure how big a role writing will play in synthesizing and communicating (since I think presentations, video, visual communication is going to play a pretty big role, too). I have a bias towards reading, but it’s definitely a changing world.

    Like

  18. The tests that came out of NCLB are different from the SATs. The state standardized tests from NCLB were designed to measure the education proficiency of an entire school district. That were never designed to track the performance of individual kids. Here in NJ, parents and schools don’t get a good enough breakdown of the scores to really provide much info about individual kids.
    Schools are unable to change themselves. So, if a school district finds that the majority of the 6th grade class doesn’t know long division, they can’t do anything about it. They can’t control factors outside the school, like family dynamics or economic pressures. They can’t fire teachers who may or may not be doing the job. They only way that they can possible react to these test scores and make adjustments is to concentrate on the top 10% of the kids who are the high test scores. They can offer them perks and place them with the best teachers.
    Do I have any quantifiable evidence that this is occurring? No, this is just a blog post that is intended to jump start a conversation. Is it possible to quantify my claims? Sure. Certainly, it could be backed with qualitative evidence, including through interviews and focus groups with school administrators and parents.

    Like

  19. Our ADD kid did well enough on the PSAT to get into dual enrollment college classes — despite having C’s in high school because she forgets to hand work in. Our son, on the other hand, has always struggled with standardized tests. I’ve actually worked for a bunch of companies that write the tests and I can tell you (and him) exactly where he fell down on SAT’s, etc. Kids lacking ‘theory of mind’ have trouble with questions like “How did Johny feel when his dog ran away,” or one memorable question that showed a woman on a street at night looking scared which was then linked to a story on street crime and dangerous neighborhoods. (My kid noticed only that she was standing hear a 7/11 and concluded that she was concerned about buying groceries, completely not noticing the other clues in the picture).
    The weird thing that happened in our kid’s school in Northern Virginia was that there were lots of people prepping their kids from the standardized tests, and so they ended up with this really bizarre cross-section of kids in their GT program. The first sorting took place in second grade based on a bunch of IQ-type tests that are apparently available on the internet if you know where to look. Basically you got all the ethnic kids who got sent to cram school and a bunch of white kids who actually couldn’t read the GT novels in third grade but who had been prepped for the test. Meanwhile, all the advanced readers got put in regular classes where they proceeded to act bored and drive everyone insane. And years later, the administrators couldn’t figure out why most of the GT kids didn’t end up in top colleges and a fair number of the average kids did.
    As long as there are tests there will always be people trying to work the system. In Fairfax county, I found out that some schools were using diagnoses and ESL measures to opt out something like 16 percent of a class’s scores from the aggregates used to rate your school. In other words, if you and your husband were both doctors and you grew up in Bombay, and you had a son who got A’s and a daughter who got B’s, they would go ahead and declare your daughter to be ESL but not your son. Uh huh.

    Like

  20. MCAS (Massachusetts) does a pretty good job of breaking down the scores. Here is a pic of my daughter’s 6th grade scores: http://imgur.com/6gICiZn
    You can see her weak point is Geometry, for example (which explains her trouble this term in 8th grade math). This kind of breakdown was phenomenally useful for E, too, as it was obvious he excelled at questions involving math or memorization of language rules, but his reading comprehension scores were weaker. This helped as we were putting together the IEP.
    You can even see how they did on each question and look up the questions they missed, though not all the questions are released.

    Like

  21. Wait a second. I thought the rap against NCLB was that the schools were throwing all their efforts into getting the lowest achieving students to meet a very mediocre minimum and ignoring higher achieving students?

    Like

  22. Wendy’s description is what our ERB scores (private school testing) look like, too. We can’t look at the actual questions (which I would love), because the test is not updated frequently enough, and we don’t have enough of a feel of the kinds of questions in each section (which we could, but no one makes this information easily available). But, I do find the tests informative in describing general areas of strength. I find the testing very useful and I’m not at all concerned about the use to which the test is put (general assessment of teaching, mostly, though it’s possible that for some kids areas of concern arise, as well), or the time spent on it (about a week).
    But, we’re a private school and we haven’t reached the stage where these data will be used to discriminate among the children. When that happens, though, I won’t be uncomfortable, because I do believe the testing is a reasonable estimate of my child’s ability (i.e. test panic or bubbling skills or something else not being tested is not significantly impacting their scores).

    Like

  23. It is possible for both “getting lowest achievers to meet a mediocre minimum” and “offering perks to the high achievers” could be happening at the same time. I hear both in our neck of the woods, and as an outsider can see evidence for both. The biggest perk the 10%ers get, though, is the opportunity to learn with the other 10%ers. That, in turn brings any benefits (including a student and parent population that has more time, treasure, and influence than the average population). They don’t get perks like smaller class sizes, though, and their access to teachers hasn’t been consistent (teachers self-select in their favor, sometimes, but not always), and, the programs get moved from location to location, disruptively (and, some think, in order to improve the NCLB scores of the building in which the program is housed).

    Like

  24. Laura, you said, Schools are unable to change themselves. So, if a school district finds that the majority of the 6th grade class doesn’t know long division, they can’t do anything about it.
    I know my kids’ former middle school used the MCAS testing to find weak areas. If formulas for finding the area of a square was a weak point for many students one year, the school emphasized instruction in that area the next year. MCAS was very good in our district in pinpointing weak grades. Students’ scores did suffer when they were in grades which had recently changed curriculum, or had teacher issues. It is not designed to be used in placement decisions, but of course, it is.
    So that group of high fliers are pampered by the school district. They are fought for by competing charter schools and private programs. During the high school orientation last week, the principal explained that we were losing about 8% of the middle school population to magnet and private schools and they were going to start recruiting kids to stay in the school system when they were in 7th grade. Losing 8% of the town’s smartest kids was a huge issue.
    About 10% of my oldest child’s 8th grade classmates were admitted to private schools. More may have applied, but only those who got in told other students. So, “at least 10%.” A large factor in applying out for those families may have been the extreme gatekeeping at the high school, limiting entry to certain AP courses. It sounds counterintuitive, but strict limits on honors/AP track enrollment can drive smart kids away. If only 7% of the strongest math students are permitted into the highest overall track, and their class rank will trump everyone else’s because they have one more AP course (weighted grading), many parents whose children have good chances of ending up in the top 10% will worry about their children’s placement.
    The Winner Take All High School: Organizational Adaptations to Educational Stratification, Paul Attewell
    (^^This happens at our public high school, the cultivation of superstars, at the expense of other students.)
    The Frog Pond Revisited: Educational Context, Class Rank, and Elite College Admission, Thomas J. Espenshade, Lauren E. Hale, Chang Y. Chung

    Like

  25. Something that I don’t often hear mentioned in these discussions is that testing itself — that is, the very act of taking a test, recalling the information under a mild stress — improves long-term retention of information. So testing isn’t only a means of measuring learning, it’s a method of learning itself. Here’s a link to a paper with the findings: http://people.duke.edu/~ab259/pubs/Butler&Roediger(2007).pdf
    To add another anecdata point, I’m not terribly exercised about testing in our affluent, suburban, midwest school district. I have kids in elementary and middle school, and I have found the state standardized test results (which are helpfully broken down in the reports sent home) to be helpful in identifying potential soft spots in their understanding which I can then address myself. My kids don’t complain about the testing; it is used as only one component of several in placement into G&T programs and tracked subjects; and it doesn’t seem to crowd out writing instruction. My 6th grader writes a three-paragraph essay each week, summarizing and explaining a scientific article she has read. I think this is a really helpful exercise; as was suggested above, I found in teaching college students that the simple ability to *summarize* an argument clearly was the foundation for all higher critical skills. I’m much more interested, at this stage, in my daughter’s comprehension and clear expression than in “critical thinking” per se.

    Like

  26. Honestly, I don’t even think that Laura meant that they losing 8% of the best test-takers. I read that the district was losing 8% of the public school students to private/magnet schools, and that this 8% was disproportionately selected from the better test-takers. I don’t seriously think that the top 8% of scorers left the public schools, but I would believe that 75% of the 8% who left might have fallen in the top 20% of standardized-test-scorers.

    Like

  27. ” If only 7% of the strongest math students are permitted into the highest overall track, and their class rank will trump everyone else’s because they have one more AP course (weighted grading), many parents whose children have good chances of ending up in the top 10% will worry about their children’s placement. ”
    This is what happens in other countries, and why there is a growing business of the next 3% coming to the US for university.

    Like

  28. “So testing isn’t only a means of measuring learning, it’s a method of learning itself. Here’s a link to a paper with the findings: http://people.duke.edu/~ab259/pubs/Butler&Roediger(2007).pdf
    I love this data, on the value of testing for retention, but it is an argument for frequent, low stakes testing, not infrequent, big-time, high stakes testing. The data supports quizzes, not week-long tests once a year.
    Also, most of the experiments on testing and retention involve declarative knowledge (i.e. content). The experiments haven’t addressed more complicated skills (to my knowledge). And, for activities like writing, frequent graded essays may be the equivalent.

    Like

  29. This is what happens in other countries, and why there is a growing business of the next 3% coming to the US for university.
    In our middle school, a large portion of the top 10% (+) applied out, depending on the parents’ views of “going private.” The top 10 to 20 percent of middle schoolers aren’t necessarily the top deciles in high school. Some students gain executive function with age, and others bloom once they get away from middle school. The rankings aren’t static. In one middle school math class, 20% of the grade was determined by math. The rest of the grade was compliance with classroom expectations.

    Like

  30. Yes, THAT. Amen @AmyP, that’s how it is in most of the country.
    Some of you are falsely equating “gifted” (an ascertainable standard) with “high-achieving” (i.e. a set of behaviors often correlated with high-SES) Profoundly gifted kids (IQ >145) are among the most at risk for underachievement – their needs are usually not served by your average US public school (except for the schools in some of these leafy suburbs you all are referencing here). The gifted research pretty much unanimously says we should have a healthy skepticism of test-scores, and only suggests using them as a diagnostic when we suspect underachievement. Radical acceleration is something we need to see more of in the less well-to-do US schools to meet these kids’ needs. See A Nation Deceived
    http://www.accelerationinstitute.org/Nation_Deceived/Get_Report.aspx

    Like

Comments are closed.