Testing, testing…

December9

DONE FOREVER from mr. nightshades photostream

Testing, it is the only objective criteria for assessment isn’t it? How can you be sure that students have “gotten” it if they can spit it back up on a test three months later? How can a test be fair if kids use aids like a calculator, how do we know that they “really” know how math if they need “help”? I could just go down the line shooting these and other “myths” away, but what takes its place, and why is that better? Okay, enough questions, let’s try for some answers, NOW!
What is the goal of education? Most of us want our students to be some blend of good and productive citizens which translates crudely into being able to hold down a job. Others may have a more beneficent vision of education, but this is the most functional goal, and testing still fails it. Let’s pull out a case study. I was discussing this recent post with my bil who is a manager at a software firm. I told him about how students in the study who showed mastery based on the formative data, did not always have mastery based on either of the summative assessments (a standardized bubble test, and a report they had to write summarizing their findings). We discussed whether mastery was in completing a task (the formative data), or passing a test. BIL thought that the written report was probably the assessment instrument that was most like work projects. Now this is where the conversation got interesting. He and his co-workers have gone to some of the best schools in the country. Many, like BIL, graduated with honors, but the biggest problem he has is that they can do the task part of their job very well (they’d pass a formative assessment), and they do well on tests (he guessed many had 800s on their Math SATs), but they can’t explain what they are doing when requests for information come from management.
My conclusion from discussions with BIL and others and based on my 7+ years in analytical work, you will rarely be given a bubble test as part of your professional life, you will be asked to solve problems, and to communicate with others. Bubble tests do not accurately assess whether someone can perform a task in context. They also are not a great way to assess how clearly folks can communicate about those tasks and projects. I hope this story illustrates both why bubble tests are not useful, but I hope that it also points the way at how we should be assessing, and why those types of assessments are called authentic. What can it look like in education? Here is an interesting article on what Nebraska tried to do. Interestingly, part of why it died was because scores on tests (NAEP) did not match how students performed on their portfolios, and it was assumed that the test was “accurate” and grade inflation was occurring in the portfolios. It never seemed to occur to folks that it might be the bubble tests that were missing something, instead of blaming the students and teachers.
I’ve blogged before about the romance of the quantifiable, and paired with that, a distrust of measures seen as subjective. Here is my suggestion, lets let Nebraska do their portfolios. Since it’s the first time we’ll be doing this under NCLB, do some audits on them. Have outsiders look at it, auditors without a stake in the outcome. They couldn’t do any worse than KPMG has lately.

Links:

by posted under politics/policy | 2 Comments »    
2 Comments to

“Testing, testing…”

  1. December 20th, 2008 at 5:10 pm      Reply PeonInChief Says:

    Actually, this one is easy. What’s tested on bubble forms is the ability to do analytic thinking, while what’s required to explain something to others is synthetic thinking. Synthetic thinking requires that you understand both your topic (they’ve got that okay, presumably) and how others will understand it. Synthetic thinking can draw connections across areas of learning to explain things–the art of thinking, as it were, rather than the science.


  2. December 20th, 2008 at 5:19 pm      Reply alicemercer Says:

    Well, it’s a sort of analysis. The problem is that since your range of answers is limited to the choices given, your analysis is limited to the answers provided.

    Now, I believe there is a place for using set choices with students. I use “dichotomy” choices with language learners and with my son who has a language delay, a lot. I don’t think you get the best assessment of a person’s thinking from this. I don’t think you get analysis. At best you get comprehension. We can do better than this.


Trackbacks/Pingbacks

  1. Mirror, Mirror… | Reflections on Teaching

Email will not be published

Website example

Your Comment:

rssrss
rssrss

Links of Interest


License

Creative Commons License
All of Ms. Mercer's work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.


Skip to toolbar