Documente Academic
Documente Profesional
Documente Cultură
THE MISPLACED
MATH STUDENT
Lost in Eighth-Grade Algebra
September 2008
by:
TOM LOVELESS
Director, Brown Center on
Education Policy
just a quarter of all students take algebra The surge continued into the next decade.
before high school.”1 The administration Eighth-grade enrollment in algebra hit 31
made enrolling all children in an algebra percent nationally in 2007, a near doubling
course by eighth grade a national goal. In a of the 1990 proportion. Today more U.S.
handbook offering advice to middle school eighth graders take algebra than any other
students on how to plan for college, U.S. math course.4 In July 2008, the State of
Secretary of Education Richard Riley urged, California decided to adopt an algebra test as
“Take algebra beginning in the eighth its eighth-grade assessment of student pro-
grade and build from there.”2 Robert Moses ficiency. The policy in effect mandates that
ratcheted up the significance of the issue all eighth graders will be enrolled in algebra
by labeling algebra “The New Civil Right,” by 2011.
thereby highlighting the social consequences At first glance, this appears to be good
of so many poor and minority students news. Transcript studies indicate that 83
taking remedial and general math courses percent of students who take geometry in
instead of algebra.3 ninth grade, most of whom completed alge-
The campaign was incredibly success- bra in eighth grade, complete calculus or
ful. Several urban school districts declared another advanced math course during high
a goal of algebra for all eighth graders. In school.5 Research also suggests that students
Are we enrolling eighth 1996, the District of Columbia led the nation who take algebra earlier rather than later
graders who know very with 53 percent of eighth graders enrolled in subsequently have higher math skills.6 These
algebra. From 1990 to 2000, national enroll- findings, however, are clouded by selection
little mathematics in ment in algebra courses soared from 16 effects—by the presence of unmeasured fac-
higher-level math classes? percent to 24 percent of all eighth graders. tors influencing who takes algebra early and
8th-Grade Total
NAEP Advanced
Jurisdiction Score Enrollment
National 281 38% who takes it late. Schools routinely assign mandating algebra for all eighth graders
Massachusetts 298 45%
incoming eighth graders to math courses mean that the nation’s students learn more
Minnesota 292 35% based on how much math students already math? Not necessarily. Although cross-
North Dakota 292 21% know. Moreover, it is no surprise that excel- sectional state test data cannot answer such
Vermont 291 26%
lent math students want to take the most a question, they can answer a different ques-
Kansas 290 39%
New Jersey 289 40% challenging math courses available to them tion: do states that enroll more students in
South Dakota 288 30% and that low-achieving students avoid these advanced math courses score higher than
Virginia 288 42% courses as long as possible. Whether algebra states enrolling fewer students in advanced
New Hampshire 288 30%
Montana 287 24%
for eighth graders is a good idea, especially courses?
Wyoming 287 32% for those who have not learned basic arith- Table 2-1 shows the 2007 eighth-
Maine 286 29% metic, cannot be concluded from existing grade National Assessment of Educational
Colorado 286 44%
evidence. Studies that test for causality, such Progress (NAEP) scores for states and juris-
Pennsylvania 286 42%
Texas 286 28%
as experiments with random assignment of dictions and the percentage of eighth graders
Maryland 286 52% students to treatment and control groups, enrolled in advanced math classes (Algebra I,
Wisconsin 286 30% have not been conducted. Geometry, and Algebra II). Massachusetts
Iowa 285 27%
The push for universal eighth-grade scores at the top (298) and has 45 percent of
DoDEA 285 40%
Indiana 285 33% algebra is based on an argument for equity, eighth graders enrolled in advanced math,
Washington 285 31% not on empirical evidence. General or more than the national average of 38 percent.
Ohio 285 35% remedial math courses tend to be curricular But several high-scoring states enroll fewer
North Carolina 284 33%
Oregon 284 39%
dead-ends, leading to more courses with students in advanced classes. North Dakota
Nebraska 284 35% the same title (for example, General Math 9, and Vermont, for example, are ranked third
Idaho 284 37% General Math 10) and no real progression and fourth in math achievement but enroll
Delaware 283 36%
in mathematical content. By completing a relatively low percentage of eighth graders
Connecticut 282 39%
South Carolina 282 41%
algebra in eighth grade—and then complet- in advanced math (21 percent and 26 per-
Utah 281 58% ing a sequence of geometry as freshmen, cent, respectively). On the other end of the
Missouri 281 33% advanced algebra as sophomores, and trigo- spectrum, the District of Columbia scores
Illinois 280 33%
nometry, math analysis, or pre-calculus as last on NAEP but continues to be one of the
New York 280 21%
Kentucky 279 34% juniors—students are able to take calculus in leaders in the percentage of students taking
Florida 277 42% the senior year of high school. Waiting until advanced math.7 The Pearson correlation
Michigan 277 38% ninth grade to take algebra makes taking coefficient, a measure of the statistical rela-
Arizona 276 32%
Rhode Island 275 41%
calculus in high school more difficult. From tionship between two variables, for NAEP
Georgia 275 49% this point of view, expanding eighth-grade score and advanced math enrollment is
Oklahoma 275 27% algebra to include all students opens up -0.09, indicating no correlation.
Tennessee 274 31%
opportunities for advancement to students Another intriguing pattern in eighth-
Arkansas 274 33%
Louisiana 272 24%
who previously had not been afforded them, grade NAEP scores emerges from examining
Nevada 271 34% in particular, students of color and from the scores of eighth graders taking advanced
California 270 59% poor families. Democratizing eighth-grade math courses. The national average in
West Virginia 270 33%
algebra promotes social justice. eighth-grade math has been rising steadily,
Hawaii 269 28%
New Mexico 268 34% increasing by 8 points from 2000 to 2007,
Alabama 266 30% Two Curious Patterns in from 273 to 281 (see Figure 2-1). But one
Mississippi 265 21% NAEP Data group stands out for not participating in the
District of 248 51%
Columbia One catch. Course-taking is a means to an score increase—eighth graders in advanced
end, not an end in itself. Students take math classes. Their NAEP scores have declined
Source: Author’s calculations from
8th grade math state main NAEP, courses to learn mathematics. Will policies from 299 in 2000 to 295 in 2007, a loss of
NAEP data explorer http://nces.ed.gov/
nationsreportcard/nde/
The Brown Center Report on American Education 3
Comparison of state and NAEP proficiency
Fig
Eighth-grade NAEP scores:
National average for students in advanced math (2000–2007) 2-1
281
280
270 273
Advanced: Algebra I,
Geometry, Algebra II
260
National 2000 2001 2002 2003 2004 2005 2006 2007
4 scale score points. The typical eighth grader information on a nationally representative
knows more math today than in 2000. But sample of 160,000 eighth graders. Unlike
the typical eighth grader in an advanced math the data used in most NAEP studies, these
course knows less. How can that happen? files require licensing for use and allow
As a cross-sectional measure of student investigators to drill down to individual
achievement, NAEP provides snapshots of student characteristics. The 2005 data are
math achievement at one point in time. The the most recent available for this type of
data cannot prove or disprove causality. But analysis. Advanced math courses are typically
NAEP data do provide rich descriptive infor- the courses that good math students take in
mation on what is going on in schools. Access the transition from middle to high school
to eighth-grade algebra has expanded dra- mathematics—in previous eras, during the
matically. Almost nothing is known about the first few years of high school. “Basic” refers
students who are taking these courses. Are we to courses taken before students enroll
enrolling eighth graders who know very little in formal algebra, including pre-algebra,
mathematics in higher-level math classes? naturally, but also general math.
One important limitation to the data.
Methods Course-taking on the eighth-grade NAEP
We tackled this question by examining is reported by students. They are asked to
data on students in advanced math courses, check off the math course in which they are
their schools, and their teachers. The data currently enrolled. Many students may not
analyzed below are from the 2005 NAEP know the actual title of their math course,
restricted-use files, providing student-level may exaggerate the level of the course, or
may for some other reason not report the courses experienced enrollment declines.
course accurately. Transcripts would Enrollment in advanced courses increased by
provide more accurate information. about 10 percentage points, from 26.7 percent
Although they have been collected on high to 36.6 percent, and in basic courses fell by
school students for other national surveys, about 16 percentage points, from 66.6 percent
transcripts generally are not available for to 50.8 percent. It appears that many stu-
eighth graders. Because of this, the NAEP dents who would have taken lower level math
data are the most authoritative in existence courses were taking algebra, geometry, or
for tracking national trends in course-taking advanced algebra in 2005. The campaign for
in eighth grade. Whatever flaws arise from algebra by eighth grade clearly succeeded in
student self-reports, there is no reason to boosting the number of American youngsters
believe that the reliability of the reports has enrolled in tougher mathematics courses.
changed significantly over time, allowing for Are all of these new students in
reasonably accurate estimates of changes in advanced courses actually good at math?
course enrollments. Unfortunately, the answer is no. In fact,
many are very poor math students, at least
What Math Courses Are Eighth as measured by their performance on the
Graders Taking? NAEP math test. Let’s consider students at
As shown in Figure 2-2, between 2000 and the 10th percentile and below—the bot-
2005 eighth graders shifted towards tougher tom 10 percent of students nationally on
courses. The percentage of students taking the NAEP test—as low-achieving or strug-
advanced courses shot up while basic math gling math students. How did their course-
2000 2005
Fig
NAEP scores of different student groups, 2005
2-3
NAEP scores
300
Low-performing eighth
291
graders in advanced
classes score even below 279
240
the average fourth-grade
student. 238
211
Source: Author’s calculations from 180
NAEP restricted-use data set: U.S.
Department of Education. Institute
of Education Sciences, National Center
for Education Statistics. NAEP 2005 120
Mathematics Restricted-Use Data Files,
Grade 8 (NCES 2007-486) and
NAEP data explorer http://nces.ed.gov/
nationsreportcard/nde/ 60
0
8th graders All 8th graders All 4th graders 8th graders
advanced classes low-achievers
advanced classes
1 Remarks by President Clinton, Education Roundtable, 6 Julia B. Smith, “Does an Extra Year Make Any 12 William Schmidt and others, “Relationship of Tracking
Springbrook High School, Silver Spring, Md., March 16, Difference? The Impact of Early Algebra on Long-term to Content Coverage and Achievement: A Study of Eighth
1998. Available at http://www.ed.gov/inits/Math/timsroun. Gains in Mathematics Attainment,” Educational Evaluation Grade Mathematics,” Michigan State University, 2008.
html. and Policy Analysis, 18 no. 2 (1996): 141-153.
13 Several years ago, a prominent education scholar gave
2 Quoted in Matthew Bowers, “Virginia and the U.S. 7 Both California and D.C. schools serve a large number a talk at Brookings in which she commented that her own
are Improving Slightly at Math, but We Lag Behind Our of students of low socioeconomic status. Nevertheless, the son had never learned the multiplication tables but went on
Economic Competitors in the Developed World,” The math scores for California and D.C. look dismal even if to graduate from an Ivy League college.
Virginian Pilot, March 28, 1997, p. B3. comparisons are made among similar groups of students in
14 U.S. Department of Education, Foundations for Success:
terms of race, ethnicity, and parental education.
3 Robert Moses, “Algebra, the New Civil Right,” in The The Final Report of the National Mathematics Advisory Panel
Algebra Initiative Colloquium, Volume II, edited by Carol 8 Other includes all other courses, along with no responses (Washington: 2008).
Lacampagne and others (U.S. Department of Education, and multiple responses, on the NAEP survey item.
15 Richard J. Murnane and Frank Levy, Teaching the New
1995), pp. 53-67.
9 In 2005, approximately 4.2 million students were Basic Skills, (New York: The Free Press, 1996).
4 Data available on the main NAEP data explorer: enrolled in eighth grade. The estimate of 120,000 comes
16 Indeed, Kennedy mentioned the first astronaut in
http://nces.ed.gov/nationsreportcard/nde/. See also Jay from [0.078 * 0.366 * 4,200,000 = 119,901.6]. All other
space, Alan Shepherd, whose suborbital flight had occurred
Matthews, “Adding Eighth Graders to the Equation,” figures in this paragraph are author’s calculations from
about three weeks before the speech. President John F.
The Washington Post, March 12, 2007, p. B1. restricted use NAEP data.
Kennedy, “Special Message to the Congress on Urgent
5 Carolyn Shettle and others, America’s High School 10 U.S. Department of Education, Foundations for National Needs,” May 25, 1961. Text available at:
Graduates: Results from the 2005 NAEP High School Transcript Success: The Final Report of the National Mathematics www.presentationhelper.co.uk/kennedy_man_on_the_
Study (Department of Education, 2007), p. 11. Other than Advisory Panel (Washington: 2008). moon_speech.htm.
calculus, advanced math is defined as pre-calculus or AP
11 William Sanders, “Teachers, Teachers, Teachers,”
statistics.
Blueprint Magazine, September 1, 1999.
ADVISORY &
REVIEW BOARD
MICHAEL W. KIRST
Stanford University
DIANE RAVITCH
New York University
BARBARA SCHNEIDER
University of Chicago
MARTIN WEST
Brown University