Standardized Tests Reveal Grading Gap
At its annual convention in July, the National Education Association approved a resolution declaring the teacher union’s opposition to “federal requirements to make significant decisions about schools, teachers, or children based primarily on test scores.” The union also voted to support legislation that would allow parents to let their children boycott standardized tests. The union further directed its lobbyists to work against the mandatory testing provisions in the education bill currently being finalized on Capitol Hill.
Although various speakers promoted the idea of using “assessments through multiple measures,” Oakland algebra teacher Judi Hirsch cut to the chase with her message to the assembled teachers: “If you want to know how your child is doing, you don’t wait seven months to get the results of a standardized test,” she declared. “You ask your child’s teacher.”
The NEA’s position fails to recognize that the main reason lawmakers have turned to standardized testing for tracking student achievement is because the grades teachers assign to student work are unreliable. (See, for example, “Best and Brightest Pull Higher Grades for Lower SATs,” School Reform News, December 2000.)
“Grade inflation is particularly extensive in high schools with a high percentage of disadvantaged students,” SchoolMatch Advisory Board Chairman M. Donald Thomas recently told a national audience of school administrators. “This indicates clearly that expectations for students are very low, and standards do not match those of testing agencies.”
As more and more state testing programs expose the gap between student test scores and the grades they are assigned by their teachers, the initial reaction of many parents and teachers is that the tests–not the grade from the teacher–must be wrong. This presents lawmakers with a difficult choice among three options:
- Develop a plan for schools to raise student achievement to the desired level of proficiency by the desired date.
- Lower the required level of proficiency.
- Put off the date by which students must reach the required level of proficiency.
For most states, the bad news is just coming in. Others already are having to make some difficult decisions.
Lawmakers Delay After More than Half Fail Math Test
After seeing the disappointing results from the high school graduation exam last year, Alaska legislators recently approved a bill to delay from 2002 to 2004 the requirement that students pass the test to graduate.
Although 75 percent of high school sophomores passed the reading test, only 48 percent passed the writing test and just 33 percent passed the math test. This year, math scores were better, but not reading and writing. Sixty-six percent of sophomores passed in reading, 46 percent in writing, and 43 percent in math.
Anchorage Daily News
June 1, 2000
D Grade Gets Diploma in High-Stakes Exam
Even though the California State Board of Education in June lowered the bar for passing the state’s new high-stakes high school exit exam from a C grade to a D grade, more than half of the state’s students still failed the test. If the state board had accepted the recommendations of a standards panel to use the C grade/70 percent cutoff, three out of four students would have failed math and one out of two would have failed English.
With the D grade/60 percent cutoff, less than 45 percent of the state’s freshmen scored high enough to get a diploma in 2004.
San Jose Mercury News
June 8, 2001
Inflationary Revision of Score-Setting
Instead of grading student performance against a set of predetermined academic standards, the Kentucky State Board of Education in June approved grading student performance on a curve.
When Kentucky parent Richard G. Innes compared the original and revised scoring results for 2000 CATS scores, he found “a massive and highly inflationary revision to the score-setting process for Kentucky’s public school assessments.” The percentage of students classified as proficient jumped from 15 to 37 percent for elementary schools, from 13 to 34 percent for middle schools, and from 21 to 29 percent for high schools.
June 12, 2001
38 Percent Fail New HS Test
Illinois high school students can’t blow off state tests any more. That’s because the American College Test (ACT) is part of the state’s new Prairie State Achievement Examination for high school students, and the results of the test will appear on their transcripts.
Even so, when the figures from the first year’s tests were released in early July, the results disappointed educators and legislators as well as students.
Forty-three percent of the state’s eleventh-graders failed in science, 41 percent failed in math, 38 percent failed in reading and in writing, and 35 percent failed in social science.
Illinois State Schools Superintendent Glenn “Max” McGee and Illinois State Board of Education Chairman Ron Gidwitz both said they had no plans to lower the pass-fail marks for individual tests in order to decrease the percentage of students who fail.
July 11, 2001
Almost Two-Thirds of Eighth-Graders Fail Math
To move to the next grade in Louisiana public schools, students have to pass the Louisiana Educational Assessment Program exam. That doesn’t bode well for the 64 percent of New Orleans’ eighth-graders who failed the math portion of the test in May, or the 47 percent who failed the English portion.
Last year, 63 percent failed math and 37 percent failed English. Students who fail the test must attend summer school and retake the test.
May 12, 2001
Baltimore Boosts Test Scores
For the second straight year, first- through fifth-grade students in Baltimore’s public schools have made significant gains on national reading and math tests. The improvement comes four years after the city and the State of Maryland formed a partnership to address the problems of what was called an “academically bankrupt system.”
Three years ago, only 29 percent of the city’s first-graders could read at or above the national average for their grade. This year, the figure is 56 percent. In math, 52 percent are at or above the national average, compared to 30 percent in 1998.
May 18, 2001
Curve Boosts Pass Rate
A month after school superintendents urged the state Board of Regents to keep the passing grade on the Regents tests at 55 until 2007, educators found the state was adjusting raw test scores on a grading curve, making it even easier for students to achieve a passing grade.
In the recent biology/living environment exam, for example, students needed to answer only 33 percent of the test material correctly to achieve a score of 55. Answering 46 percent of the material correctly would bring a student’s score up to 65, which is what the passing grade was until 1996.
The Buffalo News
June 12 and July 18, 2001
Lawmakers Bring Back Social Promotion
A few years ago, Ohio lawmakers approved a “reading guarantee” that required school districts, effective next year, to hold back fourth-graders who hadn’t passed the state proficiency test in reading. But for the past four years, the percentage of Ohio fourth-graders who passed the reading test has fallen from 68 percent in 1998 to 57 percent this year.
Faced with the prospect of having 40 percent of fourth-graders repeating the grade, the legislature abandoned the reading guarantee and allowed social promotion to continue.
The Plain Dealer
May 31, 2001
A sample question from the 1998 Ohio ninth-grade math proficiency test is: About how long is a new, standard-sized pencil? (A) 7 inches, (B) 7 pounds, (C) 7 yards, or (D) 7 ounces.
“Such a question would be more appropriate for the second grade,” notes C. Bradley Thompson, associate professor of history and political science at Ashland University. “If, as the teachers say, they are spending all of their time ‘teaching to the test,’ one wonders how long it takes to teach a 14-year-old the length of a pencil.”
Buckeye Policy Perspective
Bar Lowered to Pass Math
Students who took the high-stakes Texas Assessment of Academic Skills exam this spring had to answer significantly fewer math questions correctly to pass the test, according to test documents from the Texas Education Agency.
Two years ago, students had to get about 70 percent of the questions correct to pass, but this year they had to get only about half correct.
TEA officials said the adjustment was made because the test was more difficult than in previous years. Critics worry the lower passing standard calls this year’s improvements into question.
The Dallas Morning News
June 9, 2001
Only 7 Percent Proficient in Math in Milwaukee
Only two out of 25 students in sixth-, seventh- or ninth-grade classes are proficient or better in math, according to the first results from the Milwaukee Public Schools’ expanded program of standardized testing. Less than one-third of ninth-graders can read at a proficient or better level, with a typical ninth-grader being two years behind in reading.
Milwaukee Journal Sentinel
June 18, 2001