Friday, September 7, 2012

14 reasons why multiple choice tests suck

John Spencer wrote a post where he outlines 14 reasons why multiple choice tests suck. Consider these points:
  1. Multiple choice is shallow. I know people try to create analytical, evaluative and creative questions. However, the medium itself is one of recall and recognition.
  2. Knowledge is often connective and deeply rooted in context. Multiple choice takes away connective thinking and puts it into empty silos, where knowledge is reduced to lowest common denominator. I want students to know information. However, I also want them to know applied information. I want them to put the information together.
  3. Multiple choice is unreliable. As long as a student can guess a correct answer, every question is somewhat suspect. I get it. Teachers can look for overall trends and take out the statistical probability of guessing. However, how does that help anyone figure out if a child needs additional support with a concept.
  4. There is rarely a chance to explain why something is true. Students should be able to articulate what they know and give a defence for why it is true.
  5. Multiple choice kills the desire to learn. It might not sound like a big deal, but every time my students take a test, they are less likely to enjoy what they learn.
  6. If you try for critical thinking, multiple choice tests become subjective. Students end up with questions like, "Which of the following best describes . . ." and the test becomes meaningless.
  7. Multiple choice does not allow for nuance, paradox or mystery.
  8. Students need to see knowledge as contextual and personal. Multiple choice tests are standardized, impersonal and void of any real context. Students internalize the idea that learning is something irrelevant.
  9. It reduces self-efficacy. Multiple choice tests fail to allow students to find information themselves and make decisions about their learning. To do so would make tests "unreliable." Too many variables. Unfortunately, life has multiple variables and "learn to be a critical thinking citizen" cannot be easily measured.
  10. Multiple choice pushes students toward a narrow, cerebral definition of learning. Multiple choice does not allow for social learning. Instead, students must prove what they know in isolation. It doesn't allow for multiple modalities or differentiation, either.
  11. Google has replaced multiple choice. In an era when knowledge is instantly available, we should be seeing whether students can find information, ask deep questions, engage with sources, curate what they find and find the bias in a source. Multiple choice doesn't allow for any of these necessary skills.
  12. It's way too easy to cheat, both for teachers and for students.
  13. Learning is often a creative endeavour. However, multiple choice tests do not allow students to be creative, divergent or innovative.
  14. Multiple choice tests are often not valid. For example, if a standard is concept standard, it often cannot be assessed in multiple options. If a standard is a process or a product, they probably should be assessed by actually doing the process and making the product. So, we're left with the assessment of skills, and only those that fit nicely into multiple options.
I also wrote a post on the folly of multiple choice tests.

What do you think?

22 comments:

  1. This is the point in education discussions when I want to put my head down and cry or just walk away. Instead, I'm going to ruefully shake my head and say "Seriously? SER-i-OUSLY? SERIOUSLY?!?!" and post a semi-ranty comment.

    We're talking about the messiest, most complicated, gooiest parts of learning - assessment and you're reducing to a list including the word "suck." I get where John is coming from - he very eloquently stated in his blog (that you didn't include here) that he was frustrated by the mis-use and over-reliance on multiple choice but I don't understand the benefits of re-posting his list without context.

    I'm feeling a bit diva-y and will happily write a list of 14 defenses of multiple choice or 14 reasons why hammers suck but I'd way rather have an open conversation about quality, balanced assessment and/or using tools appropriately.

    ReplyDelete
  2. Because I'm now coming down off my rant - feel compelled to clarify that in no way am I defending bad tests, bad policy, or multiple choice abuse. Rather, I suspect our entry point should be balanced assessment systems, a focus on authentic performance tasks with authentic audiences, and validity - matching the assessments we craft or offer students to the standards. Looking forward to the conversation.

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete
  4. @Jenn, Let's pretend balance is what we want. Do we have a balance now?

    ReplyDelete
  5. In the system? Nah. In many classrooms, yes. But eliminating one means of assessment isn't going to create balance in the universe. Multiple choice tests aren't a panacea - they aren't the solution to problems as many believe, but nor is their existence the cause of all education problems.

    And speaking of balance, why is shallow a bad thing? The bottom of Bloom's is just as worthy as the top. Again and again, the issues come up with people rely exclusively on one type of assessment over others.

    ReplyDelete
    Replies
    1. I would say that there may be equal validity in the different levels of Bloom's however the objective is to inhabit the top not the bottom. So testing that forces us to dwell in the bottom is not what we should be focusing on.

      Delete
    2. I would say that there may be equal validity in the different levels of Bloom's however the objective is to inhabit the top not the bottom. So testing that forces us to dwell in the bottom is not what we should be focusing on.

      Delete
  6. This comment has been removed by the author.

    ReplyDelete
  7. This comment has been removed by the author.

    ReplyDelete
  8. This comment has been removed by the author.

    ReplyDelete
  9. This comment has been removed by the author.

    ReplyDelete
  10. Projects and performances that are in a context and for a purpose allow us to assess all levels of Bloom's in a more authentic manner than MC.

    ReplyDelete
    Replies
    1. True.

      And sticking a 16-year-old in a car and putting them on the highway is a more authentic way to assess their understanding of traffic signs and signals, but we still require them take an MC test first to ensure they know the sign for merge versus yield.

      And pushing a diver backwards off a boat into the ocean is a more authentic way to to assess their ability to use the signs for "need help", but we still require new divers to take a MC test first to ensure they can match the sign with the message.

      And putting a car accident victim in front of a new EMT is a more authentic way to assess their ability to select the right medication for a patient under stress, but we still require they take a multiple choice test where they need to select from four similar sounding mediation names to ensure they can ID the right one when seconds matter.

      Etc. Etc.

      Multiple choice will never beat performance tasks for authenticity. Nor will they measure creativity, originality, kindness, etc. That's not their job. There job is to provide a quick, convenient, way to provide evidence to the learner and teacher regarding a learner's ability to recognize basic knowledge. Multiple choice should not be the system, but in some horrible cases, it has become the only. That's bad. Horribly bad. That assessment illiteracy. That's having 10,000 spoons when all you need is a knife. It's not the spoon's fault. Nor the guy eating soup.

      Delete
    2. It's a good thing that immersing children in the realities of developmentally inappropriate real world activities or subjecting them to soul-killing, drill & kill, fill in the blank worksheets and bubble tests are not are only two choices...

      Delete
    3. Hold the phone... "soul killing?" Are you telling me the zombie apocalypse has started and I missed it? Darn. I was so looking forward to that.

      Your argument was that MC and authenticity were incompatible. I gave three examples of how they work together to support a learner. You interpretation seems to be that I'm suggesting we throw 5 year olds off the back of boats or immerse them in worksheets. This isn't an if this, than that situation.

      Reducing multiple choice, yes. Eliminating... why? Because you say they're bad? (And if your reason for eliminating them are the 14 items above, most of those need little more than a push to crumble.)

      To go back to your response, is it your plan to tell your daughter to walk out of class if her teachers ask her to take multiple choice tests in order to protect her soul?

      Delete
    4. You only have to talk to a few students about their experience with multiple choice tests and quizzes to find out how uninspiring they really are. Or, we might just need to think back on our own experiences to remember how pointless they felt.

      I would be very hesitant to minimize how damaging multiple choice tests are to a child's education.

      Delete
    5. A multiple choice tests isn't damaging unto itself. Ask any 16 year old who has passed the test for their permit. Passing a test may very well rank as one of the best days of the life. Point being - bad policy is bad. Using multiple choice as a default rather than thinking strategically through documenting evidence of student learning, providing feedback, and the why and how of of assessment is a problem. I cannot see how the test format itself is rarely, if ever, the problem.

      Delete
  11. This article did what I imagine Joe set out to do: create a vital conversation about something important. What lets us know children are learning? I gave up on tests, for the most part, several years ago. We do a lot of authentic assessment in my classroom. It works. Children can apply their learning in visible ways that are not revealed effectively by any form of test.

    We rarely, if ever, test adults the way we test children. The product in the workplace and one's ability to do the job is the test; at least in theory. When we say we are preparing children for 21st Century workplaces, we need to reflect this in their learning, the use of various technologies in effective ways, and what we, including our students, use to assess their learning.

    ReplyDelete
    Replies
    1. Agreed and I think 'rarely' is a key word. There is a word of difference between: "Multiple choice tests are being wildly overused. They should be used strategically, purposefully, and rarely. Instead of relying them on the default assessment, consider whether or not they are the best fit to what you what to capture about student learning" and "multiple choice tests suck".

      Joe has claimed there's nothing a multiple choice test can tell him he can't find in an authentic assessment. While this may be true, I maintain my examples above. There are times when multiple choice is safer, more reasonable, and more efficient.

      Where I struggle, is Joe's implied argument that somehow education/learning is such a special, special concept that we dare not quantify a moment of it. (This is not defending over-testing or bad policy. Bad policy is bad.) Numbers are used to inform decision-making and feedback in the fields of medicine and public safety on a daily basis. I struggle to see why the act of quantifying some aspects of teaching and learning is inherently bad; why collecting evidence of student learning that is discrete and lower level is, by itself, bad. Again, said with 1000 caveats. We trust that our professional trained and informed doctor won't put us on chemo based on one wonky blood test but we still expect them to collect quantifiable evidence of health. While diagnosing health issues and teaching and learning are different processes, in this example, it comes down to how the professionals use the information provided.

      Delete
  12. Good Day Joe,
    I enjoyed your article on why multiple choice tests suck. I am inclined to agree with you on many points. I recently watched a video entitled "Sir Ken Robinson: Do Schools Kill Creativity?" (http://www.youtube.com/watch?v=iG9CE55wbtY) that you may find interesting. While multiple choice tests are not mentioned specifically, I found that your article shared several sentiments with the lecture by Sir Ken Robinson. I agree with many of your points and enjoyed the conversations in your comments.
    I am currently pursuing a second bachelor's degree and plan to teach after graduation. I have been inspired by several of my college professors to diligently seek information on effective methods to avoid obsolete teaching techniques with the changing times. I find articles like yours to be very helpful in those pursuits. Thank you for your insight!
    PS. Is that Chicago Bears jersey in your photo? I happen to be a pretty big Bears fan myself!

    ReplyDelete
  13. The problem is if standardized tests don't change then teachers will always give multiple choice. That and colleges will have to change too, especially in the math and sciences. We cannot not teach mutilpe choice if we know colleges and beyond will fall suspect to such practices.

    ReplyDelete
  14. check out http://cae.org/cwra to see how different a performance based assessment can be. Content driven,summative, often muitiple choice, types do not accomplish the goal of measuring learning or the learner.

    ReplyDelete