Student comprehension is tough to judge for teachers at the helm of a packed classroom, so researchers at the University of Massachusetts are developing a program that can gauge whether students are bored, frustrated, or motivated during computer-based exercises.

UMass researchers received a grant of $890,419 this month from the National Center for Education Research to advance technology that uses sensors to detect student emotions, allowing teachers to tailor their lessons more easily around classroom victories and struggles.

"It allows them to see how their students are doing and to see what their weakest areas are," said Ivon M. Arroyo, a research scientist at the University of Massachusetts at Amherst who has developed intelligent tutoring programs for the last decade. "[The technology] serves as an assessment tool for the teachers and informs them of how each student is handling [his or her] work."

Researchers said they have overseen test runs of the emotion detectors in recent years, but they weren't sure when the technology would be unveiled in K-12 schools. The computer-based tutors developed by UMass researchers and their peers at Arizona State University help teach algebra and geometry to high school students but eventually will be available for every subject.

The tutoring program uses sensors placed in a student's seat, in the computer mouse, and on a student's wrist to detect arousal through skin conductance, a common measure for stress response. Conductance gives researchers a clear picture of the subject's nervous-system activity. The program also will use cameras to detect smiles and facial expressions that connote negative feelings, such as anxiety or frustration. Once these reactions are recorded, Arroyo and her colleagues will match each reaction with the proper emotion, giving an accurate readout for teachers.

"We're trying to make the computers smarter so they can understand the students," she said.

Beverly Woolf, a computer science researcher at the university who has developed tutoring programs for more than 20 years, said the ability to monitor students' emotional reactions to class work could be invaluable for teachers. A frustrated student isn't likely to comprehend the day's lesson, she explained.

"Emotion and cognitive functions are strongly correlated," Woolf said. "So if you improve the social intelligence of the computer, students respond the way they would to another person. Sensors allow the computer to identify students who pay attention and those too tired or bored to learn."

Don Knezek, CEO of the International Society for Technology in Education, a nonprofit organization aiming to improve instruction through the use of technology, said the tutoring program would best be used in classrooms where one-on-one instruction was rare or nonexistent.

"It is pretty evident that such a system would be extremely useful in learning situations where there might not be a human available, with both adequate skills and enough high-quality personal contact, to sense these emotions of boredom and frustration in students in their early stages," Knezek said. "Considering the role boredom and frustration play in students disengaging, and the role meaningful engagement plays in high-yield learning, interventions that can reduce the occurrence of these clearly have positive potential."