on Percentiles

Tests. We all hate them, but at the same time we know that they are a necessary thing in life. As children, we are first introduced to the idea of tests in school, as they are used as a means to measure the level of learning in a particular subject. And as we all discovered as we got older, tests are not limited to just school, but something we will have to take from time to time for the rest of our lives. Most tests are designed to reveal how well we are learning as individuals, so we know areas in which we excel, and where we need to improve. Others merely measure how well we are doing when compared to the knowledge level of others, within a set group. When I was in school this was called grading on a curve. Now it is called grading by percentile.

Most state and national tests grade using the latter version. For some reason we measure our children not on their individual knowledge, but on how they compare with other students at their same grade level. For example, Iowa uses a proficiency standard of the 41st percentile when reporting on how well students and schools are doing in areas such as reading and math. What that means is that an individual student’s test score must be higher than 40% of their fellow students at their same grade level in order to be deemed proficient. So instead of deciding proficiency based on how many questions were answered correctly on a test, the Iowa Department of Education merely groups children together and ranks them based on how many correct answers they made when compared to their peers.

Unfortunately, like grading on a curve, this form of testing obscures the actual achievement of individual students, schools, states and ultimately, our nation. It reveals nothing about the acquisition of knowledge made in school. Very few industries outside the public school system use the percentile ranking to measure skill level; you either meet their needed basic individual level of knowledge or you don’t. How well other employees do says nothing about how capable you are at your job. So why do the public schools systems use percentiles instead of net raw scoring? I think it’s for the same reason why students cheat: to avoid hard work and still give the appearance of doing well.

In 2003 the state of Iowa adopted the National Percentile Rank (NPR) metric to report cut-scores in order to determine which students are not proficient, which are proficient and which are advanced. Unlike most other states or international schools who use this ranking, Iowa chose one of the lowest proficiency metrics in the country. The state adopted the NPRs of 1 to 40 as not proficient, 41 to 89 as proficient, and 90 and above as advanced proficiency. Not only did the choice to use low NPRs instead of net raw scoring artificially boost the appearance of our student proficiency nationally, but for some reason Iowa includes the scores of those children in districts who are educated in private or homeschool environments, which effectively skews the proficiency scoring of our schools even more. Also, most other states use the 65th percentile to measure grade level proficiency (internationally the 95th is used); by Iowa using the 41st as our NPR we have effectively deemed proficiency at 1-2 grade levels lower than actual student grade level.

Because of low NPRs, and the diluting of scores by including children educated outside the public school realm, we have no idea how well our kids are actually doing. For example, in our district last year, according to the Iowa NPRs, nearly 87% of our 11th graders scored proficient in math, and nearly 84% scored proficient in reading. Initially these numbers look good, however remember that proficiency is being set beginning at the 41st percentile (i.e. grade on a curve), and not on individual student raw test scores. To give further clarity, being at a 41st percentile at the 11th grade translates to about a 9th grade reading and math level. And unfortunately, since we don’t know the range of those students in the percentile rankings above 41, we cannot determine how many of our 11th graders actually tested at their grade level (a NPR of 65%) in reading and math. Sadly, the only thing we can determine is how many 11th grade students were NOT even at the 9th grade level of knowledge, in our district’s case 13% & 16%, respectively.

We need to stop cheating, be honest, and start measuring achievement as the real world does, based on individual, quantifiable knowledge acquisition, not on how well one does when compared to their peers. We also must stop lowering the bar on how we define educational excellence, as it only makes excellence subjective, changing, and ultimately, quite meaningless. And we must stop diluting the reports with student subjects who are not truly part of the public school district. I feel that we are on a fast track race to the bottom, in search of the lowest common denominator. And for what? Just so we adults can feel good about the decisions we make? Avoid being embarrassed? No. It is time to stop the grand cover-up. It’s time to face some ugly truths on how our kids are truly doing in school, and return to more realistic test measurements of student knowledge, without any political, feel good, or CYA obscuring techniques.  No amount of new curricula, mandates, teacher development or money will improve education until we do. In order to succeed we first must know the true depth of our failure. It will be hard pill to swallow; Iowa used to be at the top of the educational game.  But for the sake of our kids, can we handle the truth?


This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply