With one set of academic standards now serving as the educational guideposts in nearly every state, questions are hovering about what the tests for those standards will look like. But gradually, details are emerging that show plans that could fundamentally change the U.S. testing landscape.
Documents issued by the two groups of states that are designing the tests show that they seek to harness the power of computers in new ways and assess skills that multiple-choice tests cannot. Those plans are very fluid, however, since several years of design, dialogue, revision, piloting, and reworking lie ahead before the assessments are ready in 2014-15. But early documents offer glimpses of the groups鈥� thinking.
鈥淭his stuff is a very big deal, and it鈥檚 a huge departure from the kinds of tests most kids currently take,鈥� said Chuck Pack, a national-board-certified math teacher at Tahlequah High School, in Tahlequah, Okla., a small town outside Tulsa.
鈥淎s classroom teachers, we鈥檙e sitting here waiting to know what our kids are going to be expected to do. We have the standards鈥攚hat they鈥檙e supposed to know鈥攂ut now how are they supposed to be able to demonstrate that? Documents like this help us get our heads around that,鈥� said Mr. Pack, who serves on an advisory board that is guiding CTB/McGraw-Hill as it designs 鈥渘ext generation鈥� assessments.
The information is trickling out in solicitations issued in the past two months by the SMARTER Balanced Assessment Consortium and the Partnership for Assessment of Readiness for College and Careers, or PARCC, for vendors to work on the tests. Those two consortia of states are using $360 million in federal Race to the Top money to create new assessments for the Common Core State Standards, which all but four states have adopted.
鈥淓very major publisher that has been a state assessment contractor in the past, and many others with an educational reform orientation, are paying attention鈥� to those solicitations, said Alan J. Theimann, the legislative counsel for the Association of Test Publishers.
A Dec. 30 solicitation by PARCC, seeking vendors to write test items, describes the consortium鈥檚 vision of its testing system in more detail than did previous documents. It expects to award that contract in April to 鈥渕ultiple鈥� vendors to design half the test items, and renew the contract to some of those vendors to craft the rest.
The solicitation covers the development of the two pieces of the test that will yield students鈥� summative scores in mathematics and English/language arts and be used for accountability purposes: a computer-based end-of-year test and a performance-based assessment given toward the end of the year. The scope of work also includes developing midyear formative assessments that are part of PARCC鈥檚 system but are optional for states.
Digging Into Text
A preliminary blueprint of PARCC鈥檚 English/language arts exam shows that the performance-based assessment, spread over two days, would involve a 鈥渞esearch simulation鈥� that asks students to read a suite of texts, including an 鈥渁nchor鈥� text such as a speech by a prominent historical figure. They would have to answer questions that require them to cite evidence from the text for their answers and write an essay. Another aspect of the performance-based test would require students to 鈥渆ngage鈥� with literature (grades 3-5) or conduct literary analysis (grades 6-11) using a combination of shorter and longer texts.
The end-of-year exam would employ six literary and informational texts and ask students to respond to machine-scorable questions, including ones that demand comparison and synthesis of the readings.
The end-of-year test in English/language arts would yield at least half of a student鈥檚 points in that topic. One-third to one-half would come from the performance-based test, according to the preliminary blueprint.
PARCC鈥檚 math test will include three types of questions: 鈥渋nnovative,鈥� machine-scorable, computer-based items; items that call for written arguments or justifications; critiques of mathematical reasoning, or proof that students 鈥渁ttended to precision鈥� in math; and items involving real-world scenarios. The performance-based assessment in math will count for 40 percent to 50 percent of a student鈥檚 points in that subject, and the end-of-course exam will yield 50 percent to 60 percent of the points.
The math exams will focus on solving problems in the 鈥渕ajor content areas鈥� at each grade level, as well as demonstrating conceptual understanding, fluency and mathematical reasoning, and applying knowledge to real-world problems.
At the high school level, PARCC will develop two series of end-of-course math tests: a traditional one鈥擜lgebra 1, geometry, and Algebra 2鈥攁nd one that integrates those topics. Those parallel pathways reflect choices educators can make about how to design math courses from the common standards.
The solicitation document answers a question that had been circulating among some educators of young children. PARCC said that its tests will be given by computer to students in grades 6-11, but those in grades 3-5 will answer questions with pencil and paper because of concerns about younger children鈥檚 keyboarding skills.
Teacher Participation
Documents issued recently by the SMARTER Balanced consortium offer a less-descriptive preview of its tests, largely because work on an earlier solicitation, to design item specifications, isn鈥檛 yet complete, and informs other parts of the test design. That request for proposals, issued in July, and its content specifications, released in August, represent the most detailed version of the consortium鈥檚 ideas. (鈥淐onsortia Flesh Out Concepts for Common Assessments,鈥� Aug. 24, 2011.)
In a request for proposals issued last month, SMARTER Balanced seeks development of 10,000 selected-response or constructed-response items and 420 performance tasks in math and English/language arts to facilitate pilot-testing in the 2012-13 school year. Most will be scored by machine, the document says.
Part of the work will be conducting research to find out which types of items are best suited to automated scoring and which must be scored by hand.
The request for proposals also asks the prospective vendor to hire and train teachers from SMARTER Balanced states to write items and tasks and review items for content alignment, accessibility, and bias. PARCC documents say that teachers will help shape the tests by serving on local committees reviewing test items. They will also be involved in developing model instructional units, diagnostic assessments, professional-development modules, and other PARCC resources.
In September, SMARTER Balanced issued a solicitation for development of guidelines for accessibility and accommodations for English learners and students with disabilities. PARCC plans such a solicitation this year, as well as requests for work on other parts of its testing system, such as its early-year diagnostic assessments and tests of speaking and listening skills.
PARCC has contracted with the Dana Center at the University of Texas at Austin to build prototype assessment tasks in math, and with the University of Pittsburgh鈥檚 Institute for Learning to generate such items in literacy. Those items are slated for release this summer.
