In Favor of Teaching to the Test
Hear that sound? It’s the joyous roar of state education agencies celebrating the U.S. Congress’ passage of the Every Student Succeeds Act (ESSA), which replaced No Child Left Behind in December 2015. Although it still requires math and reading/language arts testing at specific grade levels, ESSA gives states substantial leeway on the selection, consequences, and use of student academic assessments. States now have a new opportunity to create authentic assessments to supplement or even replace traditional standardized tests.
Authentic assessment was the reform darling of the 1990s—at least until President Bush signed the test-heavy No Child Left Behind law in 2001. In contrast to standardized tests, which typically feature multiple-choice questions and are administered once a year, authentic assessments give a fuller picture of a student’s performance. Advocates argue that just as a fitness program measures program success with a treadmill test (an authentic assessment that demonstrates a client’s fitness level) rather than a multiple-choice test (a poor proxy for measuring a client’s fitness level), schools should ideally measure student learning through tasks that:
are realistically contextualized, such as creating a plan for a community garden;
require judgement and innovation, such as creating an environmental awareness poster;
ask students to “do” the subject, such as writing a mathematical proof from start to finish;
replicate key challenging situations in which adults are truly “tested” in the workplace, in civic life, and in personal life, such as calculating the net cost of various potential colleges;
assess the student’s ability to efficiently and effectively use a repertoire of knowledge and skill to negotiate a complex and multistage challenge, such as programming a computer game; and/or
allow appropriate opportunities to rehearse, practice, consult resources, and get feedback on and refine performances and products, such as writing a lab report.
(From Wiggins and McTighe, 2005.)
Most standardized tests are what education researchers call “high inference” assessments, where you have to make a big leap between the evaluation task (e.g., a multiple-choice question) and the actual learning goals being measured (e.g., understanding chemical changes). Foes of these tests often decry “teaching to the test,” because they say these tests are poor proxy measures of what students actually know.
On the other hand, authentic assessments are “low inference” assessments because the evaluation task (e.g., a rubric assessing a student’s creation of a novel experiment to demonstrate chemical changes) is closely aligned with the learning goals (e.g., understanding chemical changes). With authentic assessments, proponents say that you want teachers to teach to the test, because the “test” assesses students as they perform complex and meaningful intellectual tasks that are in themselves learning experiences.
Yes, authentic assessments have caveats. They can be more expensive, time-consuming, subjective, and may require longer class periods as well as faculty and principals skilled in implementing the assessments. Some critics might say authentic assessments are less challenging than rigorous standardized tests, which could create lower expectations for students. The consequence, critics might say, could be students that are not sufficiently prepared for college, and the creation of a larger achievement gap between low-income and high-income students. Implemented poorly, authentic assessments are an ineffective yardstick to compare different school systems.
But there are remedies for these potential issues. School systems could pilot authentic assessments on a smaller scale, validating them by correlating results to performance on standardized tests. Schools can roll out authentic assessments to elementary school students , yet still require typical standardized tests for high school, where it is more developmentally appropriate. Or schools might randomly select some students to participate in authentic assessments while others took standardized tests, using the results of both types of assessments to better mark their progress in meeting achievement goals.
Schools can also get buy-in by touting the ideas that authentic assessments a) are in themselves learning opportunities for students—in contrast to multiple-choice tests, which appear to offer little instructional value; b) can raise standards by requiring students to demonstrate complex skills and higher-order thinking; and c) may allow vulnerable populations to incorporate knowledge and expertise that is not always valued in traditional standardized tests. Additionally, it is possible to develop and implement authentic assessments in standardized ways that ensure reliability of scoring across teachers and schools and quantify the results against normative benchmarks for student academic progress.
This column in the Washington Post notes that typical state standardized tests are “very narrow measures of what kids could do, and imbuing [traditional standardized tests] with such importance is an insult to authentic assessment.” The notion of using authentic assessments on a large scale is not new; selected schools in New York and New Hampshire have been piloting authentic assessments in lieu of standardized tests for several years. Moreover, ESSA has authorized a performance-based assessment (a type of authentic assessment) pilot program in seven states.
