Standardized tests have become unfashionable. In 2020, when the Covid-19 pandemic prompted many tests to be canceled outright, many breathed a sigh of relief. Good riddance, right?

Well… maybe. It is important to acknowledge how extraordinarily successful these tests have been. This lesson explores how tests are designed, how they work, and how they are used. It reviews the most important tests in use in California's education system, and summarizes conflicting points of view about changing them.

Standardized tests quickly measure students’ knowledge and skills in ways that allow for comparisons and support decisions. Because testing often works as a way of measuring skills, its history extends back to pharaohs, emperors, generals, and captains of industry. Because testing doesn’t always work, and because it has often been misused, that history is somewhat checkered.

Testing for Economic Success

As discussed in Ed100 Lesson 1.3, an education system that works well drives economic success — not just for the individual, but for the whole economy. That's a pretty powerful reason for an education system to use tests.

Individuals have a related, but narrower reason to care about tests. Students and parents obsess over (especially the SAT / ACT) because they are so powerfully connected to college access, and thus to job choices.

Historically, standardized tests have been used as a tool for identifying the best-prepared candidates for jobs and opportunities — and for efficiently "weeding out" the rest. In the last few decades, however, the purpose of testing has broadened. Among other things, test results serve as diagnostic tools for school leaders and teachers to choose how to spend their time. They also serve to identify where students are falling behind — a necessary step to overcome wishful thinking and do something about it.

California’s testing system has changed

California's battery of standardized tests has changed over time. From about 2002 until 2013, the foundation of the state’s testing system was the STAR tests, also known as the California Standards Tests or CSTs. These were classic one-size-fits-all fill-in-the-bubble tests based on California's state standards.

In response to the 2008 recession, the federal government provided significant funding to sustain public education. It also provided some incentive funding: specifically, through the competitive Race to the Top grant program, states were encouraged to do something long talked about: revise, harmonize and improve their grade level educational standards. In 2014 many states, including California, transitioned to a set of standards called the Common Core.

New standards required new tests. Because good tests are difficult to develop, states banded together. California and many western states joined the Smarter Balanced Assessment Consortium (SBAC); the other large consortium is known as PARCC. Not all states are officially in a consortium. In 2018 the Hewlett Foundation commissioned a report, Whatever Happened to All Those New & Better State Tests, to take stock of how the dust had settled. The review concluded that the PARCC tests are less rote than SBAC, but both are substantially better than the tests used in non-consortium states.

Sources: Illustration from Education First, High-Quality Assessment Project; research cited in chart includes two studies by the RAND Corporation on quality of state tests, AP and other common summative assessments (2011 and 2012), as well as research by the Fordham Institute (2016) and HumRRO (2016). Click image for more.

California's tests, collectively known as the California Assessment of Student Progress and Performance (CAASPP), are similar to those used in other SBAC member states. The CAASPP tests are administered on computers, and they adapt to the skills of each test-taker in order to take less time to determine a score.

The CAASPP system of tests includes tests in English and Math as well as the California Science Test (CAST), which debuted in 2017.

Which test
can best
assess
success?

Critics of standardized testing argue, in part, that it is a poor use of time. Some students find the tests dull, though computer-based adaptive tests are shorter and less vulnerable to this problem. Some argue that standardized testing is a degrading intrusion that squeezes the soul out of learning and teaching.

High Stakes Tests or Low Stakes Tests?

Standardized state tests are sometimes described as high-stakes tests. This description implies a sense of risk to students if they do poorly. In fact, students have no great reason to cheat on the CAASPP tests. Quite the contrary: it can be difficult to persuade students even to bother taking state tests seriously.

Teachers and administrators have slightly more reason to feel pressure than students do, if only for the sake of reputation. The growing focus on test scores has placed pressure on teachers and administrators that led to cheating scandals among the adults. If educators or school leaders do not believe that tests are a valid measure of learning, incentives based on test scores can create temptations.

What are the CAASPP tests?

California's Smarter Balanced tests, the CAASPP, are administered at the end of the school year in grades 3-8 and 11

California's state assessments are delivered toward the end of the school year in English and math for grades 3 to 8 and for 11th grade. (Curious what the tests look like? Try it!)

In early summer, students and their parents receive reports that show their scores. The reports visually indicate whether the student's score is near grade level expectations, above it or below it, and by how much. Here's a part of a 2019 sample report for Language Arts for a fictitious fifth grader:

The report includes prior year scores, which can help understand whether the student is generally keeping pace with grade level expectations, falling behind, or advancing.

Find much more information about the CAASPP tests here: startingsmarter.org. These tests are used as part of the California School Dashboard, which we introduce in Ed100 Lesson 9.7.

Adaptive Tests: the Goldilocks Ideal

The CAASPP tests are 'adaptive.' Administered online, they present each student with questions of varying difficulty to assess their level in as few questions as possible.

In the past, students were all given the same test, on paper. Inevitably, the questions would bore some students while frustrating others.

Administering tests on computers was a big change. It enabled the consortium to design tests that adapt to the skills of the test-taker. An adaptive test like CAASPP presents each student with a different version of the test, drawing questions that vary in difficulty (and score value). Ideally, each student takes a "Goldilocks" version of the test. Not too hard, not too easy: just right. This allows the test to more successfully evaluate each student's "level" in fewer questions and with higher certainty than with paper tests.

What are Alternate Assessments?

To meet the needs of students with disabilities, California also provides computer-based "alternate assessments" (CAA) aligned with the Common Core Standards. The goal: Ensure that students with significant cognitive disabilities achieve increasingly higher academic outcomes and leave high school ready for post-secondary options. (These assessments replaced the CAPA tests.) Watch the video to learn more. (Spanish version here).

Despite critics, tests are here to stay

Critics such as education commentator Diane Ravitch argue that standardized tests take too much time and lead schools to focus too narrowly on the subject areas that are tested. (In education jargon this is known as “narrowing the curriculum.”) Class time for art (see Lesson 6.8 The Arts), music, and even science has been squeezed because teachers and school leaders feel pressure to ensure that students score well on tests of math and English.

Annual testing for all students is a key priority of many civil rights organizations. By requiring universal participation, these tests reinforce the idea that education has to work for all students. Universal participation also reduces the temptation for schools to omit the scores of students that are less prepared.

It is legal for parents to "opt out" of California's state-administered tests. Fortunately, very few do so. Access to some federal funding is contingent on the participation of at least 95% of students. The rationale for requiring this high participation rate is to avoid the temptation for schools or districts to “hide” students that might not score well.

Public support for annual testing remains strong and broad. In surveys, about two-thirds of respondents support annual testing. About half of teachers support it, too. In response to the criticism that testing has focused only on math and English, California added assessments in science and undertook new changes in standards for the visual, performing and media arts.

Goodbye, High School Exit Exam (CAHSEE)

For many years, California required high school students to pass the California High School Exit Exam (CAHSEE) to qualify for a diploma. A one-size-fits-all bubble test, CAHSEE was very easy for most students. Still, for some it was the one thing that held them back from graduating. In 2015, the exam was suspended, and in 2017 it was abandoned for good.

The effects of exit exams have been extensively researched. For some students, there is evidence that CAHSEE served its intended role as a kick in the pants. But this effect was outweighed by its unintended role as a catalyst for students to give up on high school entirely.

Hello, CAASPP!

When high school students take the CAASPP test, they have a personal reason to care about their score. Through the Early Assessment Program (EAP), scores on CAASPP provide colleges with evidence of students' readiness for college-level work without taking remedial courses. (For more, see Lesson 9.4.)

The Real High-Stakes Tests: SAT and ACT

Will a student do well in college? If you could only know one fact about the student to predict his or her college fate, what should it be? Grade point average? SAT score? ACT score?

The best predictor of a student's success in college is his or her grade point average in high school...

According to research from the College Board (the maker of the SAT) the best single predictor of a student's success in college is not his or her SAT score. It's his or her grade point average in high school. The makers of the ACT exam agree, and so does independent research.

So why not punt the test and rely on grades alone? Because the combination works even better. Colleges give significant weight to standardized test scores in admitting students because they help predict student success in college.

Grade point averages and SAT/ACT test scores don't measure the same things. Some marvelous students are horrible test-takers. Some marvelous test-takers are horrible students. Succeeding on a test requires overcoming anxiety and distraction, and being able to quickly shift from one question to the next. Succeeding in a class requires an ability to apply oneself to deliver academic work.

...but the prediction is even better when combined with standardized test scores.

Students can significantly improve their SAT/ACT scores with practice and help. Money makes a difference: Test-prep coaches can help them focus on questions they miss and understand the material more deeply. Psychiatrists can help them learn to manage their stress and anxiety. It works, and families with money spend big bucks to give their kids every shot at success — a topic explored in depth by Paul Tough in his book The Years That Matter Most.

SAT/ACT scores can have an influence on college access in ways that aren't obvious. In a self-sorting effect, some students who actually might be quite successful at a "stretch" college don't risk applying if they think their scores are too low, or that they can't afford the test. By requiring the SAT or ACT as part of the application, colleges lose some applicants. Fee waivers are available for both the SAT and the ACT for students who qualify for free or reduced-priced meals.

Because of the unequal impact these tests can have on college access, in 2020 the academic senate of the University of California convened a task force to study whether to get rid of them. To the surprise of many, the task force recommended keeping them.

The COVID-19 pandemic created unequal barriers to SAT and ACT testing. With fewer testing sites and limited seating, students from lower-income areas faced particular barriers to test, retest, and potentially improve their scores. The National Association for College Admission and Counseling (NACAC) has called for colleges to reexamine their policies in the wake of the pandemic and beyond.

Other measures of student performance

As a broader, more engaging alternative to standardized testing, some have proposed that students should demonstrate their competency by producing a “portfolio” — the high school equivalent of an undergraduate thesis project. Students negotiate a research topic with advisors, and produce multi-disciplinary work that demonstrates their knowledge and abilities in several subject areas. Their portfolios are graded against a set of "rubrics" that incorporate state standards. The state of Vermont went farthest in formally adopting this strategy, and in the process demonstrated many of its challenges. Portfolio assessment remains uncommon, and the term has fallen somewhat out of favor. If your child is working on a "senior project" or a "senior thesis" this may be an echo of the portfolio idea.

Class Grades: Rubrics are not cubes

Standardized tests aren't the only thing that matter, of course. Most of the work that students do in class or as homework is assigned and graded by a teacher. How do teachers express what counts, and how do they decide on grades?

A rubric is a tool that teachers use to explain what they expect from students when they make an assignment. Rubrics are usually organized into columns that describe what counts. For example, a report might be worth 100 points total: up to 50 points for the quality of the research, 30 for the quality of the writing and 20 for the quality of the presentation. The rows of the rubric describe expectations for different levels of quality — for example what sort of research is most important, and what length should the paper be?

Most teachers create rubrics in a word processing program, distributing them to students on paper or as PDF documents. Online rubric tools exist, too, some of them not entirely terrible. Creating a good rubric is not easy; it requires significant forethought to clearly explain what you expect from students. Teachers: if you have a favorite tool, please leave a comment.

Of course, the results that really matter are long-term, and have little to do with scores. Real success in education is determined by the life path options that a student is able to access as a result of that education. Those options are the subject of the next lesson.

Next Steps

Parent leaders (ahem, such as Ed100 participants and PTA leaders) play a vital role in demystifying these tests. Work with your principal to organize a meeting at your school to go over the new tests and the California standards. A good resource to start with is this infographic, which answers frequently asked questions.

One of the best ways to understand a test is to try it out yourself. Here are some practice tests:

• CAASPP
• SAT
• ACT

Updated Sep 2017
Oct 2017
March 2019
Oct 2019
Feb 2020
April 2020
December 2020
June 2021