Fed up with standardized tests May 19, 2012
Posted by mareserinitatis in education, gifted, teaching, Uncategorized.Tags: education, gifted, gifted education, standardized exams, teaching
trackback
When I was a kid, I remember taking Iowa Basics tests every couple years. I remember this because I was both stunned and disappointed. I was usually impressed because my grade equivalencies placed me at least three grades ahead of then current placement with the gap widening as time went on. The disappointment was because nothing ever came of it. I sort of assumed that everyone I was going to school with must have similar scores because I was kept with the same people, in the same grade, without even so much an acknowledgement.
Well, okay, there was an acknowledgement – there were usually comments about how my math computation scores were so much lower than everything else. (This is what led me to believe, for many years, that I was bad at math.)
My kids haven’t used Iowa Basics, and I find this very disappointing. In a move that I can only assume is a result of No Child Left Behind (or, as I affectionately like to refer to it, the “Lake Wobegon Law” because everyone must be above average), there has been a shift away from tests like Stanford Achievement or Iowa Basics to NWEA Map testing.
The only way I can describe this is useless info that’s providing a moving target. The test provides percentiles and approximate ranges for competencies in various subfields. It is frequently renormed. In many respects, it’s the same as any other standardized test.
My beef is that, as far as I can tell, the only purpose of the test is to see how your student(s) compare with the rest of your district or nationally. On the other hand, I will say that it’s not the only one that does this. However, it seems like there are a lot of schools moving this way, and I see it as a huge detriment. The reason is that I don’t think you can make decisions about a child based strictly on their performance compared to a norm. However, that’s exactly what teachers want to do. They see an area of relative weakness in a child and want to hold them to that level for all of their abilities. I am left to ponder why it is they never want the child to be working at the level where they are capable and make an attempt to bring the weak areas up to par with the strong areas. Of course, if you have nothing to determine where they’re actually achieving, it’s hard to implement that type of education.
This leads me to wonder: how does a child working at age level help them to develop skills above age level? If you’re teaching a child stuff s/he already knows, aren’t you just holding them back?
The complaints I received about my ‘lousy’ math computation scores are one example of this. I have several tests showing this problem which constantly elicited comments from teachers about how I was poor at math. I get the impression that they looked for personal weaknesses but never really made the connection that my average was different than most of the other kids. Their solution, therefore, was to have me work on more computation at grade level.
Scores that only consist of a percentage relative to norms tell you is that one’s performance relative to everyone else may be an area of weakness. It doesn’t tell you, however, where you’re really achieving. It’s a bit different if you have a grade equivalency sitting next to the norms. It turns out that my ‘lousy’ math computation scores implied that my computation was equivalent to the average child two grades ahead of me. And it should be fairly obvious that if they wanted to me to be achieving more strongly in computation, they would have been giving me more computation at 2-3 years ahead of grade level. Unfortunately, that’s not what happened, and most often, it’s still not. It’s a lot harder to dismiss a child’s achievements when you have a solid basis of comparison (a kid two or three years older) than some vague percentile. Those percentiles don’t give teachers a true picture of achievement; how many teachers have frequency tables for a normal distribution sitting nearby? My impression is that it leaves them only feeling that when a child is at a very high level, the child is learning and thriving in their current environment. They have the mistaken impression that the child is having their needs met, when in reality, the child could be seriously underperforming relative to their potential. Likewise, they may get the impression that a child is struggling but fail to realize that it’s because they lack basics from prior years.
I therefore would like tests to go back to giving grade equivalencies. I think this illuminates the level of child achievement and gives teachers a better idea of what they are actually dealing with. There is a good amount of research showing that teachers are actually some of the worst identifiers of children’s intellectual gifts, and taking away the frame of reference that grade equivalencies provide is going to make it worse for the child and parents or other advocates.
I also blog at Engineer Blogs, home away from home to some of the best engineering blogs.
At our mid-year conference the first grade teacher told us DC’s grade equivalencies in all of his subjects, but that wasn’t from a standardized test. I’m not actually sure where they came from. They also had his grade equivalencies for his second placement test before he started kindergarten (which we were able to tell to the crappy private school we looked at, and realized they had no idea what to do with him). As far as I can tell, they haven’t gotten the state standardized tests back yet for the year.
My sister was denied the gifted program one year in middle school because her score in something on the ITBS dropped from something like 99% to 97% and since we lived in a university down she didn’t make the cut-off that year. My mom, of course, complained, and, they reinstated her before the year started.
Depending on the test, you can get grade equivalencies by looking at the norms, so it may have been from the test.
My parents went round with the TAG program in the little town where I started school. I was not the ‘typical’ gifted kid, primarily because my parents were poor and didn’t have the resources to do all the ‘extras’ that were expected as part of the program. Also, I resented that the TAG program was piling more work on top of my other classes. (I still had to make up homework on the days I was at the TAG program.) Teacher kept trying to kick me out the 2 years I was in the program. When we moved to a bigger town, the teacher wanted to recommend me for the gifted program there, but based on prior experience, my mom said no because it was the same sort of pull-out program.
:(
My nickname for No Child Left Behind is “No Child Gets Ahead”. What it really does is ensure that gifted and creative children are not encouraged to acheive their potential as they should be, because so much energy must be focused on bringing underacheiving kids up to the required levels. It’s one of the greatest disservices we have ever done to our children and it’s an embarrassment that it still exists.
I recently heard a program on NPR about how countries with the highest standardized testing scores have the lowest levels of creativity and innovation. What we are really doing is teaching kids not to think outside the box because the goal is to think like everybody else.
I think this is the link but I can’t listen to it from my mobile to confirm so I’m not 100% sure: http://www.npr.org/2011/04/28/135142895/ravitch-standardized-testing-undermines-teaching
It’s completely ironic that other countries are trying to move to a more “American” style of educational system, while we’re moving toward the Asian style (which definitely is more rote learning and based less on creativity). I think a lot of this is because education is being viewed as training for a job and not training for life/thinking.
I actually have hope for the use of computerized adaptive tests, (like NWEA’s MAP test) in that perhaps the value-added advocates will have a real basis for providing appropriate curricula. As it is currently, if all of the test items are within two grade levels of the nominal grade, the chances of a single missed test item drastically effecting the score at the highest (or lowest) end of the curve are fairly high. With CA tests, they can at least get an appropriate baseline – provided the tests go high enough.
Years ago, I used Accelerated Reader’s STAR tests to explore how my GT students were doing. But, even with this computerized adaptive test, it was hard to see students’ progress, since the test didn’t go high enough for most of my older students.
The stupid thing is that it’s really easy to figure out the equivalencies based on the norms, so I don’t understand why they just don’t print them on the results.
You’re right: the adaptive tests should go high enough. MAPS testing is supposed to go through grade 11 using the same test, as I understand it. According to their online literature, tests should be accurate up to scores of 245, so as long as they aren’t within about 5 points or so of that, there’s hypothetically no way to hit a ceiling.
We got DC’s Metropolitan achievement test results for first grade back yesterday. They were odd. He got some problems wrong, but the weird part was that for the sections where he didn’t get anything wrong, the percentile and stanines were maxed out in the lower 90s and number 8 respectively. DH wondered what’s the point of having a test that maxes out at such a low level for that grade– is it more precise for numbers closer to the mean? I thought it might be because it’s only 1st grade and they can’t ask too many questions, but the results say most of the sections have 30-40 questions, which seems like it should be a reasonable number.
From what I understand, the tests are much more accurate toward the middle of the curve and can drastically change at the extremes from just a single missed question. This is actually an argument for computerized adaptive tests and grade equivalencies. Then, if the child is getting all of the questions correct, the computer gives him/her harder questions, until s/he is missing questions. With a standard test, 10 percent of the students will get a score of 90% or above. With a CAT, you can see the huge differences in that 10%. Some will be able to go substantially higher. Interestingly, the CATs don’t have to be longer to achieve this result. There is an algorithm for skipping questions that turns out to be pretty good at pinpointing the child’s level – provided that the test has a high enough ceiling.
It’s hard to see a CAT for first graders though. I guess really it’s hard to see the point of standardized testing in general for first graders. It makes sense to have the thorough individual testing that’s more expensive if a kid is suspected of being at a tail and needs additional accommodations, but otherwise for little kids I have to wonder what the point is. (Of course, we didn’t start standardized testing until 3rd grade when I was growing up so perhaps I’m biased.)