UMass Amherst Computer Scientists Develop an Emotion-Sensitive, Computer-based Tutor That Improves G

[FarCenter]

UMass Amherst Computer Scientists Develop an Emotion-Sensitive, Computer-based Tutor That Improves Girls’ Math Scores

March 10, 2010
http://www.umass.edu/newsoffice/newsreleases/articles/101058.php
Contact: Janet Lathrop
413/545-0444


AMHERST, Mass. – As schools gear up for the Massachusetts Comprehensive Assessment System (MCAS) statewide testing of mathematics skills in grades 3–8 and geometry in grade 10 this spring, some students in Holyoke, Turners Falls and Deerfield will be helping to fine-tune a computer-based, emotionally perceptive math tutoring software developed by University of Massachusetts Amherst computer scientist Beverly Woolf, Ivon Arroyo and colleagues, which can improve standardized test scores.

In earlier studies conducted in Greenfield and other schools, the software has improved student math test scores by 10 percent, a critical difference for those who are struggling to pass. As Woolf explains, “Our original work was to find out where girls needed extra attention and how to give it to them. According to our studies, the extra support they need compared to boys is more about emotion than information.”

In April, in another round of studies supported by the National Science Foundation, Woolf and colleagues will offer the tutorials to schools in Holyoke and Springfield. Students in study groups of about 100 are offered white, black and Hispanic learning coaches by the software, to enhance attractiveness. “We want to improve students’ relationship with math early, which can be so important to their career choices. Once you close off math, you close off most of the sciences, as well,” says Woolf. To prevent that, the program virtually assigns an individual aide to each student.

Woolf and Arroyo know from previous work that girls in fifth grade thrive on extra attention, and respond well to supportive characters and positive feedback. So they developed computer-based tutorials that use such techniques. Most recently, they’ve added sensors and cameras so the computer can recognize when students are happy or stressed, fidgeting, frustrated or feeling confident. Guided by such cues, the “learning companion” character reaches out with encouraging words to praise a student’s effort, offer a hint or suggest that trying again is an important aspect of learning.

As the computer scientist explains, “Girls get equally good grades and express interest in math and science at the same level as boys in elementary and middle school, but by the time they get to high school they’re expressing more frustration and dislike of math and they start to do worse on achievement tests. So years ago we began to explore how to keep girls’ interest in math and science alive, and keep their test scores up.”

Interestingly, boys mute or turn off the learning companion character twice as often as girls when working through the mathematics tutorials. “Gender does seem to matter for learning mathematics at this age,” comments Woolf. Girls appreciate the emotional support that matches their mood that is given by the computer software, but it seems less important to boys. “Our position is that everyone can learn, it’s just a matter of how each one learns best. It’s very important in education to find the key to student success,” she adds.

At present, the software is correct about 70 to 80 percent of the time when using its camera to detect when the student user is confident and happy or bored, anxious or frustrated. Other variables that cue the software to a student’s emotional state include the time taken to answer questions; number of hints requested and grip on the mouse. In their current studies, the researchers are searching for the optimum combination of sensors and camera data to best predict a learner’s frame of mind.

To anchor the software in real-world problems, Woolf, Arroyo and colleagues developed several diverse characters in a community called “Wayang Outpost,” based on a real NSF-sponsored team of biologists working in Borneo studying endangered orangutans. Many of the math and geometry problems encountered in the 45-minute tutorials are linked to real problems at Wayang, such as figuring the total area of roofing material needed to build a new jungle research station.

For Massachusetts students threatened by failure on math and geometry tests, the test score improvements achieved by using Wayang Outpost emotion-sensitive tutorials to reinforce lessons in the weeks leading up to MCAS tests are extremely valuable, Woolf says.

“Not only do the characters in Wayang Outpost help many students feel better about their math skills, they lead to more time spent on the problems, and more enjoyable time. So they really learn the concepts being presented in each module at their own pace. Our characters are helping groups who need special attention with emotional or cognitive needs to spend that extra time that leads to more complete learning.”

Other groups, such as low achievement and special needs students have also benefited from this software. Special needs students have academic, physical, social or emotional needs and might need a classroom aide, repetition or extra time to learn well.