Tests adapt to students; will Utah adapt?
Richfield • No one says a word in the computer lab.
But when the fourth-graders return to their classroom after taking the tests, their curiosity explodes.
"When we get back to class, everyone's like, 'How did you do? How did you do?' " said Abby Peterson, a fourth-grader at Pahvant Elementary in Richfield.
It's not the kind of excitement that school testing typically inspires. But computer-adaptive tests, such as those given in the Sevier School District in south-central Utah, are not traditional fill-in-the-bubble tests. These tests adapt to each student's ability level, deliver immediate results and help teachers identify a student's strengths and weaknesses, proponents say.
And within four years, all Utah students will likely take them.
Utah is part of a consortium of 31 states that recently won $176 million in Race to the Top federal stimulus dollars to develop and implement computer adaptive tests by 2014-2015 as part of a new testing system. They will likely replace Criterion-Referenced Tests (CRTs), now given toward the end of each school year. And they will reflect new Common Core State Standards, which will change what students are expected to learn in each grade in math and language arts.
The consortium has not yet decided which company will develop the tests, but the principles behind them will likely be similar to those used in Sevier and some other Utah districts, created by the Northwest Evaluation Association.
"It's the direction that the whole country is moving, and it's the direction that testing across all our schools should be moving," said state Superintendent Larry Shumway. "These will be tests that will provide much better information to students, parents and teachers about how students are doing."
A change to teaching • On a recent school day, each of Lani Baker's eighth-graders at Red Hills Middle School took a reading test, but no pencils were in sight.
And no two tests, given by computer, were exactly the same.
One boy answered a multiple-choice question asking him to choose the word, based on its Latin root, that fit a given definition. He chose the correct answer, so the test gave him a harder question next. Had he answered incorrectly, he would have gotten an easier one. That's a big change from CRTs.
When the boy finished his test, his score popped onto his screen and Baker's.
Baker said she has mixed feelings about the test. It takes time out of the school year and doesn't paint a full picture of a student. But on the other hand, she said, it's the best test she has ever given.
"I love what this test can do," Baker said. "No other test I've ever seen gives us this kind of information. It's useful."
In Sevier, all students in kindergarten through 11th grade take the tests in the fall, winter and spring. District officials say the assessments allow teachers to measure each student's progress and adjust instruction accordingly. That's also a change from CRTs, which measure whether students hit certain performance levels not their individual progress over one school year and whose results typically aren't revealed for days.
"I prefer the adaptive testing because you can look at growth," said Tammy Anderson, a third-grade teacher at Pahvant. "You can see where a child started the year and then ended."
Also, students can score below or beyond their grade levels.
"Right now with adaptive tests, you get to see what level a student is on, whereas before when you were testing you were just seeing whether they were on level or not," said Terry Christensen, a Red Hills math teacher.
A corner of Christensen's classroom is a testament to that specificity. The corner is covered with laminated charts that show which skills students need to review, focus on and which they're ready to learn, based on their scores.
"It's really helped us change the way we teach," said Pahvant Principal Selena Terry.
The tests also give parents more specific information, said Cade Douglas, Sevier director of student learning.
"I used to go to parent-teacher conferences and didn't have a clue what was going on," said Nan Singleton, who has a fourth-grader at Pahvant and older children in other district schools. "Now I understand how my child is doing in each area."
Singleton now receives a progress report for her daughter three times a year with graphs showing how she compares with other students in the district and nationwide in math, reading and language usage. It shows where she stands in 13 areas within those subjects.
It's information that goes far beyond a letter on a report card.
Questions remain • Though Sevier parents and teachers have been positive about their experiences with the tests, questions remain about the adaptive assessments and those that will be developed by the consortium.
Will the new testing eat up too much computer lab time, which could otherwise be used for instruction?
How will Utah, which is already cash-strapped, pay to implement the tests?
Are schools putting too much emphasis on testing?
The new system the consortium is developing will consist of computer-adaptive tests given up to twice a year. It will also include optional interim assessments and tools for teachers to improve formative assessments, which are informal tests and quizzes teachers use in their normal instruction.
Myron Mickelsen, Sevier superintendent, said through careful planning, computer labs can be made available for testing without compromising instruction. And Judy Park, state associate superintendent, said it's still unclear how Utah will pay to implement the testing, though the state has a few options.
Mickelsen also said adaptive testing is time well spent. Instead of taking time away from instruction like CRTs, the adaptive tests are an integral part of instruction, he said.
"We certainly didn't start looking at adaptive testing because we wanted to increase the amount of tests," Mickelsen said. "We just couldn't estimate any kind of accurate growth measure with the tests we had."
Scott Marion, associate director of the National Center for the Improvement of Education Assessment, said adaptive testing has advantages and disadvantages. The tests enable teachers to more efficiently test the full range of student achievement in a given subject and make cheating more difficult. But it can be difficult for the tests to provide diagnostic information. And results regarding specific skills within a subject aren't always reliable because students answer only a few questions pertaining to each skill area, he said.
Douglas acknowledges the tests aren't true diagnostic tools, but said they're better than others. He said the tests are more about screening and measuring where individual students stand at different points in time.
Overall, Mickelsen said it's difficult to say whether the district has made significant academic improvement since it started implementing the tests more than three years ago. It's difficult to compare students' achievement on the CRTs to their achievement on the adaptive tests, and the district hasn't been giving the test long enough to cite a big trend one way or another. He said the district hopes to soon use the data to help develop and determine effective classroom instruction.
But teachers, parents and administrators say they're seeing change on an individual basis. Douglas said he sees the difference both as a district administrator and a parent with a second-grader and fifth-grader.
"I know right where my kids are and where I want them to be at the end of the year," Douglas said. "Now I as a parent can help in the whole process."
It's the future of education in this country, he said, and Utah just happens to be at the forefront.
What are computer-adaptive tests?
The assessments change to fit students' ability levels as they take them, helping teachers identify students' strengths and weaknesses. If a student answers a question correctly, the next question will be more difficult. If a student answers a question incorrectly, the next question is easier. Sevier is one of several districts and charter schools that have been giving adaptive tests for several years. The feds so far have denied Utah's request to allow those districts to give adaptive tests in place of CRTs, but have awarded $176 million to a consortium of states including Utah to develop new adaptive tests to be used in 2014-2015.