Under the guise of reform, Chicago’s schools are unfairly punishing thousands of innocent children whose only crime is posting a reading or math score deemed unacceptable for promotion. Somehow according to this policy, which has gained favorable national attention, failing huge numbers of children annually will solve the problems of public education. It won’t.
As an elementary school principal in Chicago, I presided over many graduations, and the importance of that ceremony to the children and their families was confirmed every spring. Thus the following headline from the Chicago Tribune of June 17, 1997, stunned me: “Tears of joy—and sorrow.” Above the headline: “For 8th graders at Von Humboldt School, Monday was bittersweet as 59 pupils graduated and 45 didn’t because their standardized test scores failed to make the grade.”
One of those 45 was Eddie, who scored a 6.8 in reading and 7.2 in math, achieving the graduation standard of 7.0 in math but just missing the 6.9 needed in reading. It was a borderline situation that the principal felt justified a waiver. The waiver wasn’t granted .
“It’s messed up,” Eddie said. While Eddie got that one right, he probably didn’t know that his chances of graduating high school had just been significantly reduced.
I feel sorry for all the Eddies of this world. They are being royally shafted by people who should know better. Scratch that. Make it, do know better. Anybody who has taken Statistics 101 or Test and Measurements 101—and that includes all professional educators—knows the limitations of norm-referenced standardized tests like the Iowa Tests of Basic Skills.
For those who haven’t taken Statistics 101, I offer this primer.
Let’s say that a group of 6,000 students has been selected as the norm group for 8th grade. Each of these students takes the test and gets a “raw” score that is equal to the number of questions he or she answered correctly. Imagine now that each student is given a placard with his raw score on it, and the children are taken to an open field. The child with the lowest score is placed on the left side of the field, and the child with the highest score on the right side of the field. The others are told to place themselves in line according to the number on their placards. Near the middle of the line, children are standing hundreds deep because their scores are very close together. But children at ends of the line are pretty lonely.
The chief test-maker then directs his lieutenants to start at either end of the line and begin counting off groups of 60 children. Each group of 60 now represents 1 percent of the tested population. In the future, when other children take the test, they will be graded according to where they would have stood in this line. Students whose raw scores put them at the 50th percentile are deemed popularly to be “at grade level,” even though the 50th percentile represents simply the middle scorers.
Test-makers also report scores by stanines, which is short for standard nine. Stanines 1, 2 and 3, which span the first 22 percentiles, are considered below average. Stanines 4, 5 and 6, which span the 23rd and 76th percentiles, are considered average. And stanines 7, 8 and 9, which span the 77th and 99th percentiles, are considered above average.
To make schools happy, test-makers also convert percentiles into grade-level scores, such as Eddie’s 6.8 in reading. The score Eddie got generally gets read as the eighth month of 6th grade. In reporting scores to parents and the public, school districts overwhelmingly choose grade levels even though percentiles are more accurate and stanines are vastly superior for instructional purposes. The grade-level system seems so accurate, but the perception is deceptive. An examination of Eddie’s scores shows why.
Eddie’s reading level of 6.8 puts him at the 24th percentile and, therefore, just inside the 4th stanine, which test-makers consider average.
Now look at his math score. On the 8th-grade math test, a 7.2 is at the 22nd percentile—2 points lower than Eddie’s reading rank—and, therefore, in the 3rd stanine, which test-makers consider below average.
So, Eddie was denied the opportunity to graduate with a 24th-percentile score in reading but deemed smart enough to graduate with a 22nd-percentile score in math. As Eddie said, things were “messed up.”
Whatever the classification scheme, though, some child will always be one question away from the next higher category, which is precisely why taking a year from the life of a child on the basis of one point on one test is nothing short of criminal.
This becomes even more apparent when the issue of test reliability is examined. As in polling, test scores have margins of error. On a test with a “reliability coefficient” of 0.95, which is highly reliable, a score of 100 could have a seven-point margin of error, meaning that a student who scores 100 could just as well have gotten a score as low as 93 or as high as 107.
That means that if Eddie had taken the reading test on another day, he might well have made the 6.9 minimum for graduation to high school.
A goodly number of the students who made the test grade after summer school likely would have done equally well simply with a retest. You can see that from the retesting at Chicago Vocational High School, where a second round of testing allowed another 89 9th-graders, half of those retested, to gain sophomore standing.
Nature—maturing four months between the May and August Iowa testing dates—probably helped some marginal students clear the test barrier. And some surely benefited from the wake-up call and the summer instructional program. In the end, though, the pass rate for the summer program wasn’t much different from the pass rate of the CVS retesting .
Forty-five percent “failed” despite small classes, individual attention and the intensive, scripted, step-by-step instructional program. They are the dirty little secret that nobody wants to talk about.
Remember our imaginary line? Most of the 45 percent would be dispiritedly shuffling into place at the far left end. They were at the bottom of their class in the spring when they were retained, and, judging by recent studies, they were at the bottom of their class when they returned to school in the fall—one year older—repeating 3rd, 6th, or 8th grade. The 45 percent are not the Eddies, who missed promotion by one-tenth of a grade equivalent. These children are lodged solidly in the first and second stanines.
There is an alternative to massive retention, and it begins with an enlightened promotion and retention policy that requires schools to justify a proposed retention to the satisfaction of the child and his or her parents. That is, the child and parents must agree that retention would be a benefit.
No child should be retained more than once and never in 8th grade. As a general rule, retention should take place in 3rd or 4th grade or in 6th or 7th. Children should be retained only after careful monitoring for two to three years. When it is determined on an individual basis that retention is in order, the school should consider moving the child back a grade in the spring, so he could be promoted with his new classmates at the end of the semester .
This, in essence, was the promotion and retention policy I followed when I was the principal in a poor neighborhood school. Under this policy, it was not necessary to retain many children; when it was, no parent objected.
Children retained under this policy tended to be socially and/or physically immature, had a history of lengthy illnesses and excessive absences from school, were new to the urban environment or had a history of frequent transfers. These mitigating factors were easily identified, and retention was usually accepted by the child and the parent as a positive educational act done in the best interests of the child. In an ideal world, the child would not go through the same experience in which he was already unsuccessful; but in the real world, you make do with available resources.
In the case of marginal 8th-graders, a special summer school should be offered that would strengthen the skills students will need in high school: reading for comprehension, outlining, note-taking, writing papers, research, problem solving and improving study habits. It should be optional; enrollment should depend on schools’ convincing parents that it would help their children get a high school diploma. This same program, suitably modified, should be available to children at the lower grades as well.
In the zeal to push Eddie’s standardized test scores up a few points, the schools are neglecting high-scoring students like his classmate Maylene, who scored 12.4 in reading and 9.9 in math. Summer programs should be offered to these bright youngsters, too, so they can see that learning is exciting. Let them read, talk, write. Forget about advanced placement. Time enough for that in high school. Allow them, for six weeks, the joy of pure learning without concern for grades or tests or rank. And again, modified accordingly, these programs should be available at the lower grades.
And finally, what about those youngsters who fall at the far left of our norming line. I’m reminded of a day long ago I when I was downtown and noticed a newsboy selling papers, making change, and greeting customers with a speed and accuracy I couldn’t match. I recognized him as one of my EMH students (Educably Mentally Handicapped). A couple years later, I saw him again in my school. I was chatting with one of my clerks when he entered the office with a package. Now a delivery truck driver, he was courteous, business-like and as efficient as he had been as a newsboy.
These are healthy, normal, feeling youngsters in every respect; except they come up short by the arbitrary standard of standardized tests. They are valuable and valued human beings, and, yes, their self-esteem is important. They don’t need a test score to tell them they are not as swift in school as their classmates. They know this everyday of their lives.
For these youngsters, I also would offer a voluntary summer program, one aimed at showing them that learning can be a pleasurable experience.