69传媒

In the face of rising opposition to testing, the PARCC consortium has decided to carve 90 minutes off its 10- to 11-hour-long assessment, and shift the start of testing to later in the school year.

The redesign of the test was approved Wednesday in a unanimous phone-conference vote by the board of governors of the Partnership for Assessment of Readiness for College and Careers, or PARCC. Eleven states and the District of Columbia belong to the consortium, which created tests aligned to the Common Core State Standards. The revised design will take effect with the 2015-16 PARCC tests in English/language arts and mathematics, which are given to students in grades 3-11.

In addition to shortening the test, the new design also folds PARCC鈥檚 two testing windows into one 30-day window, and requires that states wait until three-quarters of the way through the school year before starting the testing period.

Currently, PARCC states have a 12-week window to give the longer, more-complex performance tasks, between Feb. 16 and May 8. They have an eight-week window for the end-of-year component, which contains shorter-answer items: from April 13 to June 5. Within those windows, states choose 20-day periods in which to administer the tests.

Next year, states will have one 30-day period to give the tests. It can鈥檛 begin until 75 percent of instruction has been completed鈥攄ay 133 in a 180-day instructional year鈥攁nd it must be finished by the time 90 percent of instruction has been completed.

This year, teachers in states that began early鈥攕uch as Ohio, which opened its testing season on Feb. 16鈥�complained that they hadn鈥檛 had enough time to teach their material before testing began.

Fewer 鈥楿nits鈥�

The redesigned test would also reduce the number of sessions that students must sit for, but it would lengthen them a little as well. Currently, students take the PARCC test in eight or nine 鈥渦nits,鈥� or sessions, depending on grade level, and those sessions vary from 60 to 90 minutes each. Next year, students will take six or seven units, depending on grade level, and sit for sessions ranging from 75 to 110 minutes each.

Materials presented to the PARCC governing board show that even as they sit for longer 鈥渦nits,鈥� or test sessions, students will spend 60 minutes less on math, and 30 minutes less in English/language arts, because there will be fewer short-answer items on the test. The number of performance tasks remains the same, but fewer questions鈥攁nd accordingly, points鈥攚ill be given in the short-answer sections of the literacy and math exams.

A small subset of students, however, would not benefit from a reduction in PARCC testing time next year. That鈥檚 because they鈥檒l be chosen by the consortium to field-test English/language arts questions for the following year. That will add 75 to 110 minutes back into the test for those students.

Like any test-maker, PARCC must continue to field-test questions for future test forms. Typically, such questions are sprinkled into each year鈥檚 operational test. But given the increasing objections of parents and policymakers to long tests, PARCC decided to confine that added time to a representative sample of 15 percent to 25 percent of its students, instead of lengthening the test for all students.

Response to Criticism

The changes in PARCC鈥檚 test design aren鈥檛 the first since it won $185 million in federal funding in 2010 to design tests for the common core. In 2011, the consortium undertook a major revision, dropping its 鈥渢hrough-course鈥� design, which would have parceled the test out into four pieces across the school year. Last year, it dropped some English/language arts questions from the test in a bid to manage its length. It also made its test of speaking-and-listening skills optional.

The other consortium that built common-core tests with federal funding, Smarter Balanced, also had to respond to criticism about its test鈥檚 length. It shortened the test by several hours by cutting back on performance tasks. Now it takes students seven to 8陆 hours, depending on grade level.

Even with the changes it鈥檚 made to its tests over time, PARCC has been grappling with pushback from state and district leaders, teachers, parents, and students because the tests take up so much time. PARCC鈥檚 assessment chief, Jeffrey Nellhaus, appeared before the Ohio legislature last month as lawmakers considered legislation to cut back on testing. He reassured the House education committee there that PARCC had heard their concerns and was working on shortening the test.

Nonetheless, the Ohio House last week passed a bill eliminating funding for PARCC.

The Colorado legislature, concerned about testing time, passed a law earlier this month that exempts 11th graders from the PARCC exams. Chicago announced it would boycott the PARCC test in 90 percent of its schools, citing concern about the test and schools鈥� technological readiness, but it later agreed to administer the test in all schools when federal officials warned the district that it could be punished for failing to test all students as required by the No Child Left Behind Act.

Wednesday, Mr. Nellhaus said that the PARCC states were trying to 鈥渞espond to the field鈥� by changing the test design.

鈥淧arents are concerned about the amount of testing time. 69传媒 were primarily concerned about the burden of having two testing windows, and having to keep the testing windows open over such a long period of time,鈥� he said in an interview.

Others said the design changes will mollify only some critics.

鈥淚t might make life easier for school administrators that have to arrange testing schedules, and for teachers who have to adjust their curriculums,鈥� said Scott F. Marion, the associate director of the Center for Assessment, a Dover, N.H.-based company that offers technical assistance to states on assessment. 鈥淏ut the folks who are opposed to testing in general, and who are against the common core because they see it as a federal overreach, the only change that will appease those folks is if PARCC disappears.鈥�

Questions Raised

Testing experts identified a couple of areas of concern about the new design.

Barbara S. Plake, the retired director of the Buros Center for Testing at the University of Nebraska, Lincoln, said that by moving the testing window later into the year, PARCC might complicate the question of comparability between this year鈥檚 tests and next year鈥檚. That鈥檚 because some students in 2014-15 took the performance tasks as early as mid-February, but next year, test-takers will have the benefit of more instruction before they tackle those items, she said.

Another area of concern for Ms. Plake, who serves on PARCC鈥檚 technical advisory committee, is the design of the sample of students who will be chosen to field-test items. If the consortium doesn鈥檛 take great care to obtain an appropriate sample of students of widely varying performance levels, it could compromise the integrity of field-testing those items, she said. She also expressed concern about the added burden of testing time on students chosen for field testing. 鈥淚t might be considered unfair to some students,鈥� she said.

By reducing the number of items on the test, the consortium runs some risk of increasing measurement error and decreasing the reliability of the test, said Robert L. Brennan, who serves on PARCC鈥檚 technical advisory committee and is the director of the Center for Advanced Studies in Measurement and Assessment at the University of Iowa.

Lori Michalic, who teaches English at Tallmadge High School in Ohio, welcomed the change in design. She had struggled to cover enough content before her district began testing in mid-March.

鈥淚t will be incredible if we can shorten the amount of testing time and not lose that instructional time,鈥� said Ms. Michalic, Ohio鈥檚 2015 teacher of the year. 鈥淭hat was the biggest complaint from our staff, that students were consistently being pulled out of class to take the tests.鈥�