In January, State Superintendent Tom Torlakson recommended $1 billion in changes to the state’s battery of tests so California’s 4 million students will have to complete tasks such as writing paragraphs and multiple-step math problems. He also recommended suspending most state tests for a year as California, along with 44 other states, moves to tests aligned with the Common Core, national standards detailing what every K-12 student should know in math and English.
“The ability to engage in critical thinking and solve complex problems cannot be reliably assessed with the kinds of multiple-choice tests that are the centerpiece of our current system,” Torlakson wrote to the legislature.
Testing Knowledge, or Activity?
Testing experts disagree. “It is a radical change as to what the test is for,” said Doug McRae, a retired test developer who has worked on tests for K-12 students across the nation, “which is to measure the results of instruction.”
Before California enacted its State Standardized Reporting (STAR) tests in the late 1990s, McRae said, educators quarreled over opposite testing approaches: whether to measure what a student has learned over a course or to give teachers periodic feedback on student learning. “The two purposes are incompatible,” he said.
Typical multiple-choice tests such as STAR measure critical thinking well, said researcher Richard Phelps, who has conducted a systematic review of all the research available on standardized tests.
“There is nothing ‘deeper’ about the constructed-response test item format,” he said.
In fact, the fewer multiple-choice questions a test has, the less reliable it is for indicating what a student knows, because multiple-choice can cover more topics faster, said Sandra Stotsky, a former Massachusetts standards and testing commissioner.
“All essay answers allows too much subjectivity in rating, especially if what is being assessed is ‘critical’ thinking,” she said. “I don’t know what is necessarily intended by that phrase, and it will vary with each topic.”
The current STAR tests will expire in July 2014. Torlakson wants to have the new computerized tests and Common Core curriculum in place by the 2014-15 school year. Some school districts have already started testing software and logistics.
“I don’t think California is anywhere near to having the hardware, bandwidth, or know-how, by 2015,” says McRae.
In Sacramento, legislators’ reactions are cautious.
“I am very supportive of moving to the Smarter Balance assessments as part of the the Common Core State Standards,” said Assembly member Joan Buchanan (D-San Ramon). “My concern is whether schools have the capability. We need to undertake a thoughtful process of evaluating what it will truly take to convert to a computerized testing system.”
A central concern for legislators is the cost, Buchanan said. Torlakson acknowledged the greater expense of developing, administering, and scoring the new tests.
The new tests will also take more time to administer. The Smarter Balanced Assessment Consortium, one of two groups developing Common Core tests, estimates an increase in testing time of 50 to 300 percent.
Less Public Information
With less information about whether students have actually learned what parents expect in each grade, the tests will reduce public knowledge of school and teacher quality, McRae said.
“The Torlakson legislation is to help the teachers,” said McRae. “There a strong anti-accountability theme.” This leads to the question of how to measure teacher performance if the new tests do not accurately indicate the level of instruction or student knowledge.
The California Teachers Association did not return requests for comment.
“We have to look at the totality of our assessment system and what we expect, as a state, from our assessments and system of accountability,” said state Sen. Carol Liu (D-La Canada Flintridge), chair of the Senate Education committee. “The tenor of the [superintendent’s] report, from my perspective, is about getting it right, not fast.”
Stotsky suggests it would be better for these new, instruction-focused tests to be administered at the local level, leaving the standardized tests intact for state accountability purposes.
Image by Renato Ganoza.