Publication Date: 2007-10-30
Monty Neill gave this talk to an emerging alliance of education, civil rights, community, parent, disability organizations that have come together to oppose a proposal that students who do not pass a state test cannot get a diploma. We can all learn from it.
Thank you for inviting me to talk with you about high school graduation tests. I am
pleased to see a diverse alliance of organizations coming together to oppose the proposed
mandated state exit exams. These tests have caused a great deal of damage to students
and schools across the nation. It will take a strong, united and effective effort to block
this proposal, and FairTest looks forward to working with you to accomplish that goal. At
the same time, we must promote positive approaches to improving education for students
whose schooling opportunities remain far too limited.
FairTest has worked with educators, parents, civil rights groups, and community
advocates in many states and at the national level to oppose the use of standardized tests
as stand-alone graduation requirements. We are a small, non-profit advocacy group,
located in Cambridge, Mass. Our website is http://www.fairtest.org.
History of graduation tests
Graduation tests were introduced in the south in the late 1970s into early 1980s. They
were the product of compromise between governors who wanted to raise taxes for
schools Ă˘€" the south clearly had the weakest education systems Ă˘€" and business leaders who
agreed to support tax increases in return for assurances of improvements. The deal took
the form of minimum competency tests (MCT). Florida was the first. Civil rights leaders
went to federal court in a case that reached the fifth circuit. The appellate court approved
the use of tests provided students were given an adequate opportunity to learn; in practice
that adequate opportunity was very weakly defined. The tests spread across states, next
appearing in northern industrial states with sizeable African American populations Ă˘€" New
York, New Jersey and Ohio, for example.
By about 1989, 23 states had or planned to have MCTs. But many never implemented
them, and some dropped them, so by 1995-96 there were 16 states with mandatory
graduation tests (Neill, 1997). The number was stable until the late 1990's when there
was a renewed push for these tests. States like Arizona and California have added them.
There are now 26 that have or say they will have mandated graduation tests. These will
affect 70% of all kids and close to 90% of racial minorities. The pattern of these tests
being more prevalent in states with more minorities remains true. The question then is,
does this benefit these students? FairTest's answer is no, and I will use a good part of this
talk to explain why this is true.
A final note on the history: many of these states have some sort of appeal or alternative
process. Most such appeal systems help very few students. New Jersey is the primary,
maybe only, exception to that point, and its alternative assessment is under attack.
Some test consequences
Tracey Newhart, a young woman with Down syndrome from Cape Cod, Massachusetts,
completed her required high school coursework and was accepted to study culinary arts at
Johnson & Wales University. But because she could not pass the state high school exit
exam, the MCAS, the college rescinded her admissions offer (Rothstein, 2003).
In Baltimore, 16-year-old Portia Dyson earns A's and B's in her coursework, works in a
University of Maryland research laboratory, and dreams of studying science and nursing.
Test anxiety Ă˘€" a well-documented phenomenon Ă˘€" caused her to fail three of the four state
exams required for graduation. Now she is must retake high school courses she already
passed, instead of the subjects she needs to succeed in college (Bowie, 2007).
Tracey and Portia are real people. They are just two of tens of thousands of students
across the nation who are Ă˘€ścollateral damageĂ˘€ť from the graduation testing explosion.
Pennsylvania should think twice before going down this road. Evidence shows Ă˘€śhighstakesĂ˘€ť
tests are the wrong prescription for what ails public education.
The ills of many public schools are undeniable. Like other states, Pennsylvania has vast
disparities in educational access, quality and outcomes. The record demonstrates,
however, that exit exams are a false solution for these very real problems. Graduation
tests simply add another punishment Ă˘€" denial of a diploma Ă˘€" to the victims of
inadequately funded education. The victims are disproportionately low-income and
minority youth, students with disabilities and English language learners. But graduation
tests are a harmful distraction from addressing real issues, a false and misleading solution
that harms children without benefiting society.
Proponents of graduation tests ignore the real consequences. Avoiding the need to
address deeply entrenched social and educational issues, they promise miracle cures. In
reality, the harmful side effects of exit exams include narrowed curriculum, teaching
reduced to little more than test prep, increased dropout rates, and profoundly demoralized
students. These damaging consequences most powerfully affect low-income and
Exit exam promoters promise narrowed achievement gaps and overall score increases.
That has not happened. The number of states with graduation tests has steadily risen
over the past two decades. During the same period, however, the National Assessment of
Educational Progress (NAEP) reports no narrowing of the achievement gap among racial
groups at the high school level (Neill, 2005). Nor have average scores increased -- there
has been no rising tide to lift all the boats. I noted that southern states generally
implemented these tests in the 1980s, but the south as a region continues to lag behind on
educational indicators such as NAEP, graduation rates, and college attendance.
Other major independent studies have examined whether high-stakes testing improves
learning outcomes by comparing states with such tests to states without them (Nichols,
Berliner & Glass, 2006). If anything, the results favor those without high-stakes tests.
The Texas Higher Education Authority reported that the need for remediation at Texas
public colleges increased after that state imposed a graduation test (Haney, 2000). The
biggest high-stakes testing operation these days is the federal No Child Left Behind law.
The rate of improvement on the National Assessment of Educational Progress has actually slowed down since NCLB was put into effect (FairTest Examiner, 2007). In short, high-stakes testing does not produce improved learning.
A major reason for the lack of progress is that high-stakes testing narrows education.
Untested subjects are ignored, while tested topics turn into test coaching programs. Test
prep is like holding a match to a thermostat and believing the room will get warmer:
Scores may rise on that test, real learning does not.
Let's talk about this issue in more detail. The key justification for the imposition of the
tests is a paper on the Pennsylvania State Board of Education (2007) website, "High
School Graduation Requirements and the 21st Century Economy." I want to debunk a few
of the claims made in that paper.
First, the general claim is that the kids must pass standardized tests so Pennsylvania and
its students can compete better in the global economy. Some evidence on this issue:
Ă˘€Â˘ Economies of post-industrial nations scoring in the middle of pack on
international comparison tests did better than those scoring highest, going back to first
international comparison in 1964 (Baker, 2007). I will get to why in a minute.
Ă˘€Â˘ On a recent international study of children's overall well-being in economically
advanced nations, the U.S. was at the bottom with England (UNICEF, 2007). Across six
categories, it did best in education (slightly below the mid-point). That is, education
relatively held its own, even bringing up other areas that are not doing well. It may well
be that other areas Ă˘€" health care, housing, community stability and such Ă˘€" are more of a problem for children's learning and futures than our schools.
All this does not mean we should ignore education. Consideration of the major purposes
of education - equality of opportunity, citizenship and lifelong learning Ă˘€" means that we
must indeed markedly improve education, particularly for low-income and minoritygroup
The second claim in the Board of Education paper is that if students pass the PSSA, they
are ready for college and less likely to need remediation. The implication is that if we
prep kids for PSSA Ă˘€" or comparable graduation tests Ă˘€" then they will be college ready.
This is a logical and statistical fallacy Ă˘€" it assumes causation when there is only
It is far more likely that students who do well in well-resourced schools, those students
who get the famous well-rounded education and one that inculcates the ability to think in
and across subjects, will do well in college and on the PSSA. It does not follow that
focusing on PSSA tests will produce high-quality learning outcomes, especially on
important things no one- or two-hour standardized paper-and-pencil test can measure.
Those outcomes are often essential to doing well in college. Examples include: extended
writing in subjects; reading complex text in subjects with strong comprehension; ability
to analyze and summarize what has been read; ability to synthesize and evaluate
knowledge, information, data; ability to engage in research, produce research reports and
successfully complete complex projects; any kind of extended work. Achieve (2005),
which supports graduation tests, surveyed employers and college teachers of first-year
students and found these are the kinds of skills actually needed. Let's not pretend the
PSSA or other state exit exams measure them.
And what about the so-called soft skills? Oral communication, teamwork, problem
solving, creativity, drive? They too are prized by employers and colleges but are driven
out of schools by the emphasis on standardized tests.
All in all, the focus on graduation tests will undermine, not improve, the quality of
education in terms of economic needs. I mentioned that the U.S. economy overall did
better than economies with higher test scores (Baker, 2007). (This of course leaves aside
the issue of how well the economy served people at the bottom, impacted the
environment, etc.) The U.S. economic gains are very likely due to greater ability of U.S.
students to problem solve, be creative, and so forth. That is not to denigrate content
knowledge, but as Einstein said, "Imagination is more important than knowledge" (Baker,
2007). We should bolster content knowledge for those students who need more content
knowledge, but not at the expense interest, curiosity, creativity and imagination, which
tend to wither with an intense focus on standardized tests.
Let me turn now to the consequences for individual students.
High-stakes testing reduces the high school graduation rate. Texas introduced exit exams
in 1992. Fifteen years later, a record 40,200 students in the Class of 2007 were denied
diplomas based on the state tests (Radcliffe and Mellon, 2007). California has seen a
dramatic decrease in graduates since it imposed a mandatory exit exam in 2006 (FairTest
Examiner, 2007). National independent research confirms a link between graduation tests
and higher dropout rates (Warren, et al., 2006). The more difficult the graduation test, the
more the dropout rate goes up (Dee & Jacob, 2006).
In 2006, BostonĂ˘€™s annual dropout rate rose sharply, from 7.7% to 9.9%. At the same time,
the city has suffered a wave of youth violence. Boston City councilors, who solicited the
views of local young people on why violence was rising, reported, Ă˘€śStudents Ă˘€Â¦
expressed massive frustration and boredom with the endless drilling and practice of the
MCAS test and test preparationĂ˘€Â¦ Far too many students describe their school experience
as an MCAS-centric environmentĂ˘€Â¦ [as a result] the incentive for students to remain in
school is tenuous" (Ross et al., 2006).
In every state, the impact is greatest for low-income and minority group students, English
language learners, and students with disabilities. Consider Massachusetts:
For the Class of 2006, statewide, the DOE reported MCAS pass rates of 97% for whites,
86% for Blacks, 87% for Hispanics, 79% for students with disabilities and 72% for LEPs
(MA DOE, 2006) But when you take into consideration the dropout rates for each group,
pass rates are probably closer to 88% for whites, 71% for blacks, 64% for Hispanics, 64%
for students with disabilities and 54% for LEP students (MA DOE, 2007).
Unable to produce evidence of real success, exit exam boosters say, "WeĂ˘€™re not doing
these students any favors if we just 'give them' a diploma." But what is gained if students
have nothing to show after playing by the rules and passing required courses for 12 years
of schooling? Students without diplomas earn much less, are far less likely to maintain
stable families, and are far more likely to end up in prison. Denying a diploma based on a
test score does neither student nor society any favors.
If exit exams really enhance equity and school quality, why are southern states -- the first
to adopt graduation tests -- still mired at the bottom by any measure of educational
performance? Why, in short, should Pennsylvania follow the failed practices of
Mississippi and Alabama?
Massachusetts, my home state, would appear to be an exception. It introduced statewide
exams and then a mandatory exit test. Exit exam supporters point out that NAEP scores
are up in grades 4 and 8, and Massachusetts leads the nation. Massachusetts also fares
well on SAT college admissions tests, and on more important measures, such as the
percentage of students who go on to college. But achievement gaps remain and the
dropout rate has increased. The fact is the state always did well on many measures, and
the gains in test scores are most likely due to a massive infusion of state educational
funding, directed mainly at low-income districts.
The truth is that race and class performance gaps reflect more on what happens outside
the classroom than inside. A recent analysis of high school test scores published in The
Connecticut Economy (Heffley, 2007) found socioeconomic factors alone account for
about 85 percent of the variation in test scores in four subjects in that state. Virtually
every study ever done finds that socio-economic status is the primary explanation for test
scores. Pennsylvania can do better than putting accountability on the backs of its children
while failing to address the underlying economic and social inequalities.
Some say exit exam unfairness can be addressed by offering the tests several times a
year, with opportunities to pass extending beyond completion of grade 12 coursework.
In fact, this is inadequate. First, every test has an error range. On a well-developed normreferenced
test, the range is about 10 points Ă˘€" e.g., a 50 could really be a 45 or a 55
(Rogosa, 2001). PSSA will not be more accurate, and taking a test twice or even six times
will only solve this problem some of the time. Second, test anxiety, a well-studied area, is
a serious factor. Estimates show that sizeable percentages of test takers suffer measurable
anxiety; for many, it significantly impairs performance (Hembree, 1988). Retake
opportunities won't solve that problem Ă˘€" indeed, having failed once, a student's anxiety
may be more likely to increase.
There are also those who claim that schools will ignore kids unless there is the pressure
of grad tests. However, the nation has the NCLB law, which demands a focus on lowerscoring
students. Unfortunately, NCLB does not produce good results in terms of school
improvement, just as graduation tests are not producing real increases in student learning.
There are better ways to go about school improvement.
The choice is not between imposing graduation tests and doing nothing to improve
education. Fixing the problem of unequal schools and inadequate outcomes requires
many actions, from ensuring funding equity for low-income cities and towns to better K-
12 programs -- including professional development and high-quality assessments -- to
having expectations of a 'well rounded' education for all children.
Pennsylvania must reorder its priorities and pursue public policies that address the
foundations of children's academic success: health care, nutrition, and living wages for
working parents, along with high-quality teachers, access to a strong curriculum, and
All that said, there are things that can be done in the realm of assessment. First, the
weight generally given to tests must be reduced to stop narrowing and dumbing down the
curriculum. Rote learning to address the kinds of questions that can be answered by
filling in bubbles will not benefit students or our society. Adding a few short written
responses won't help much. Instead, assessment must be reconfigured in three ways.
First, develop and use real diagnostic or formative assessments. Research shows
formative assessment has perhaps the most powerful positive effect on learning of
anything that schools can directly control, and the results are more powerful for lowperforming
students (Black and William, 1998). But it is rarely used, and teachers are not
well-skilled in this area. These days, we are seeing districts use so-called "benchmark" or
"interim" tests that look like the big tests. These are to real formative assessment as
multiple choice tests are to real assessment Ă˘€" a weak imitation. Implementing formative
assessments will require time and money, professional development and time and space
for teachers to work together.
Second, implement a statewide system that uses multiple forms of assessment, mostly
under local control. Nebraska does this Ă˘€" it has a quality control system, but the
assessments are local (Roschewski et al., 2006). They are not all excellent, but they are
getting better, and in the process teachers become better teachers. In my handouts is a
two-page summary of what we propose in Massachusetts, the CARE plan (1999). Ironically, the Massachusetts law called for a comprehensive system of assessments. That is what the state started to do, until politics intervened and we got a single state test. Our proposal would provide all the information needed for real evaluation and accountability.
Third, think anew about what students should be required to achieve before they earn a
diploma. Ensure the resources are there to enable students to meet those goals. Then think
about how students can demonstrate this learning, what are the various ways they can do
it, and how the state can check up on the system. The CARE proposal provides one way
to do that. In any event, there is absolutely no need to impose a one-size-fits-all
graduation test to answer any of these questions. As I have explained today, the real fact
is that graduation tests hurt, not help, students, schools and society.
These are recommendations for improving assessment and using assessment to improve
schools and student learning. They are not more general recommendations for improving
education. I think that starts with a combination of adequate resources and powerful,
ongoing, educator-controlled professional development. Certainly a strong curriculum is
essential, and viable options for different students and healthy relations between schools
and parents are also vital. But we either get serious about systemic, long-term positive
change in schools, or we succumb to fraudulent quick fixes that do more harm than good,
such as graduation tests.
What is necessary then is a powerful movement among educators, civil rights groups,
religious denominations, community organizations and businesspeople to stop the
proposal for Pennsylvania to impose a high-stakes graduation test. I think you can do
that. While you do that, you should work as well for real solutions to the actual deep
problems facing education in Pennsylvania.
Monty Neill is Co-Executive Director of FairTest, the National Center for Fair & Open
Testing. Lisa Guisbond, Testing Reform Analyst at FairTest, contributed to this report.
Monty can be reached at firstname.lastname@example.org. FairTest's website.
Achieve. 2005. Rising to the challenge: Are high school graduates prepared for college
and work? http://www.achieve.org/files/pollreport_0.pdf
Baker, K. 2007. Are international tests worth anything? Phi Delta Kappan, October; V89,
Black, P., & Wiliam, D. 1998. Inside the black box: Raising standards through classroom
assessment. Phi Delta Kappan, V80, N2 (October): 139-144, 146-148.
Bowie, L. 2007. Test anxiety can put diploma in jeopardy, Baltimore Sun (September 19)
Coalition for Authentic Reform in Education. 1999. A call for an authentic state-wide
assessment system. http://www.fairtest.org/care/accountability.html.
Dee, T.S. & Jacob, B.A. 2006. Do high school exit exams influence educational
attainment or labor market performance? Social Science Research Network, April.
FairTest Examiner. 2007a. NAEP results show children still left behind under NCLB.
FairTest Examiner. 2007b. Exit exam update: WA, TX, CA, AZ, MA. January.
Haney, W. 2000. Ă˘€śThe Myth of the Texas Miracle in Education,Ă˘€ť Education Policy
Analysis Archives. V8, N41. http://epaa.asu.edu/epaa/v8n41/ and
Heffley, E. 2007. What do CAPT scores really tell us? The Connecticut Economy,
Hembree, R. 1988. Correlates, causes, effects, and treatment of test anxiety. Review of
Educational Research, V58, N1, 1988.
Massachusetts Department of Education (2006). Progress Report on Students Attaining
the Competency Determination Statewide and by School and District: Classes of 2006
and 2007 Malden, MA: MA DOE. June.
Massachusetts Department of Education (2007). High School Dropouts 2005-06
Massachusetts Public Schools. June.
Neill, M. 2005. School Beat: Notes on the recent NAEP test results. Beyond Chron.
(August 4). http://quartz.he.net/~beyondch/news/index.php?itemid=339.
Neill, M. 1997. Testing our children: A report card on state assessment systems.
Cambridge, MA: FairTest.
Nichols, S. L., Glass, G. V, & Berliner, D. C. 2006. High-stakes testing and student
achievement: Does accountability pressure increase student learning? Education Policy
Analysis Archives, V14, N1. http://epaa.asu.edu/epaa/v14n1/.
Pennsylvania State Board of Education. 2007. High School Graduation Requirements and
the 21st Century Economy.
Radcliffe, J. & Mellon, E. 2007. TAKS tests cost 40,000 Texas seniors chance to
graduate, Houston Chronicle (May 12)
Rogosa, D. 2001. Shoe Shopping and the Reliability Coefficient, Educational
Assessment, V7, N 4: 255-258
Roschewski, P., Isernhagen, J., & Dappen L. 2006. Nebraska STARS: Achieving Results.
Phi Delta Kappan, V87, N6 (February): 433-437.
Ross, M. et al. 2006. Report of the Special Committee on Youth
Violent Crime Prevention: Working Together to Increase the Peace (June): 10.
Rothstein, K. 2003. MCAS puts aspiring chef in a pickle, Boston Herald (July 17)
UNICEF. 2007. Child poverty in perspective: An overview of child well-being in rich
countries, Innocenti Report Card 7. UNICEF Innocenti Research Centre, Florence.
Warren, J.R., Kulick, R.B., & Jenkins, K.N. 2006. High school exit examinations and
state-level completion and GED rates,