When Research Projects Replace State Tests

The cafeteria at the Institute for Collaborative Education looked like a science fair on the day students took their equivalent of a Regents science test. Teens crowded around lunch tables displaying their projects with poster boards and graphs. Teams of evaluators moved among the students, hearing about research and experiments conducted over the past few months. 

This Manhattan school is one of 48 with a waiver from the state to offer alternatives to most of the five Regents tests required to graduate. Students still must take the English exam but for the others they can provide portfolios or special projects.

At I.C.E., eleventh graders take neuroscience in addition to biology and chemistry. They study the brain and conduct experiments on their fellow students, often using words and images. A few kids were allowed to work with cockroaches donated by a local college. One project looked at how teenagers respond to different colors and sounds. Another pupil studied laughter, and a few students explored racial bias.

Asha Reed-Jones showed middle and high school students at I.C.E. pictures of black and white people, and tested their reactions to different positive and negative sounding words such as "marvelous," "joyful," "painful" and "awful."

"Typically, if you took longer to associate black-American with good than you did black-American with bad, it showed you had more of a bias that was negative toward African Americans," she explained to her team of five evaluators. They included an I.C.E. teacher and a graduate student in neuroscience from New York University, along with three of her fellow students.

The evaluators spent 20 minutes asking about her conclusions and methodology, such as how she tracked the age groups.

Principal Peter Karp said the teams of evaluators were matched carefully to each student. They may include parents, outside experts and peers although teachers had the final say. They considered whether students' work supported the research and whether they clearly stated the purpose and hypothesis of the experiment. There is also a written component that's graded solely by teachers. In the end, there are no numbers or letters; students must meet or exceed standards to pass.

"Students very often do not meet standards in the first round of the evaluation," Karp said. "In fact I would say approximately it's between 10 and 20 percent of students have to extend their work into our summer program."

Ann Cook, executive director of the New York Performance Standards Consortium, pointed to data showing the schools that use these alternate assessments have higher graduation and college acceptance rates than the citywide average. She said this included students who struggled, and minority groups.

Students at I.C.E. were true believers. After learning that she had achieved a grade of honors for her presentation, the highest possible score, Asha Reed-Jones thanked her evaluators and squealed with relief. She had been working on her racial bias project since the fall. She acknowledged a Regents test might have been easier, but it wouldn't have been as meaningful.

"I feel like if students really did this kind of project they would put a lot more effort into what they were doing," she said, adding that unlike a test, you can't just study the night before.

 While there is a lot of interest in this alternative to high-stakes testing, it can take years for the state's education department to approve a school that wants to join the consortium. And, Principal Karp added, this approach might not be right for every school because it demands so much individual attention from teachers, and smaller classes.