The Proper Use of End-of-Course Exams in Determining High School Graduation

States and districts should not bar students from graduating based solely on standardized test scores. The Standards on Educational and Psychological Testing of the American Psychological Association, American Educational Research Association and National Council on Measurement in Education states that a major decision about a student should not be made "on the basis of a single test score." Dozens of educational and civil rights organizations join FairTest in endorsing this basic principle and opposing high-stakes graduation exams.

Exit exams generally do not improve student learning, and they tend to narrow the curriculum. They also have devastating effects on many young people, particularly on students of color, children from low-income families, the disabled and English language learners. (See FairTest fact sheets on graduation tests at

Unfortunately, 26 states, containing the great majority of the nation's students, have or plan to impose exit exams. Some states, however, are moving from a single "comprehensive" test to subject-specific end-of-course exams.

Subject-specific tests could do a better job than general tests in assessing a worthwhile variety of knowledge and skills, but states should not use them as sole hurdles for determining high school graduation. Doing so would violate the Standards' injunction against making high-stakes decisions solely on a test score and perpetuate the harmful consequences for curriculum and instruction. That approach also would block the possibilities of developing innovative ways to teach and assess.

With proper use, end-of-course tests could be a fairer and more educationally sound process in states that now insist on some form of exit exam. The following recommendations envision end-of-course exams as part of a larger assessment system:

• State-mandated end-of-course exams should count for perhaps 15% of the course grade in a few key subjects. They should not be "sink-or-swim" barriers to graduation.

• The subject-specific exams should assess thinking and doing, reasoning, creating, and problem solving, as well as basic information and routine procedural knowledge. Thus, they will need to include performance tasks in addition to having a modest proportion of multiple-choice and short-answer items.

• The exams can be scored locally, with spot-checking to confirm accuracy, as is the case for the New York State Regents exams and other nations' assessment practices. This approach can address the need to return results quickly while including complex performance components. States can build systems to review samples of answer sheets to ensure quality and accuracy in scoring. Teams of teachers from other schools might do the initial scoring or rescoring to ensure independent oversight.

• States, acting individually or collaboratively, should create banks of good assessment tasks that teachers can use at their discretion. Use of cognitively complex tasks will support teaching higher order skills. Their use also will facilitate teachers' use of projects in the instructional process, prepare students for exams with performance components, and help build the basis for an overall stronger assessment system.

• States should allow schools and districts to use high-quality alternatives to end-of-course exams. For example, the New York Performance Standards Consortium won the right to substitute performance assessment tasks in lieu of four out of five high stakes Regents exams. This approach to assessment leads to innovative curricular design and teaching. An expert body should approve such assessments. Studies of the consequences of such alternatives, including success in college, should be conducted, as the Consortium schools have done. (See

• The performance tasks and school/district alternatives can together form the basis for a system in which state exams become optional once schools and districts are capable of compiling validated bodies of evidence of student learning. In such circumstances, limited testing could be one check on the system.

• As a fundamental principle, graduation requirements should follow the provision of adequate opportunity to learn the material on which students will be assessed.

Developing such a system would address the need to assess higher order and complex skills and knowledge that have proven impossible or too expensive to assess with standardized exams –those now often referred to as "21st century skills." Constructing a system along these lines also would allow development of alternative curricula as well as assessment that could be valuable to future educational improvement. States should not stifle the development of superior approaches in the name of standardization.

States that do not have exit exams should not implement them. Instead, they should help districts strengthen their assessment practices and build banks of tasks and projects. They also should overhaul assessment practices across the grades to rely primarily on local and classroom-based evidence of student learning.

Such systems are technically feasible, as shown in the Expert Panel on Assessment report, Assessment and Accountability for Improving Schools and Learning: Principles and Recommendations for Federal Law and State and Local Systems (at They are common in other educationally strong nations. Politically, there is growing recognition that current systems have many harmful consequences while failing to assess too many important aspects of learning. Thus, the ground is set for states to recreate their assessment systems, including graduation requirements.

• For more detailed information on equity and quality in assessment, see Principles and Indicators for Student Assessment Systems; at 



alt end of course plan - final-ltrhd.pdf17.31 KB