“Hi, my name is Amanda. I like to discuss poetry and politics over coffee.”
“Way too high. Is there any way to adjust the pitch without changing the speed and duration?”
“Hi, my name is Amanda…”
“Huh. Way too low now, but the speed didn’t change at all. Go with it.”
Meet “Amanda.” Amanda is one of many synthesized voices currently being used in an experiment conducted by Omosola Odetunde ’13, Emily Song ’13 and Lorin Dole, a first-year doctoral student in communication. Their research question: if a robot gradually changes its voice to match the voice of the user, will the user like the robot better? This question could be applied to everything from self-driving vehicles to kitchen appliances.
Their class, Communication 168/268, started by Clifford Nass in 2002, explores a wide spectrum of questions relating to human-computer interactions, such as: does multitasking while listening to music help one be more efficient? Is it possible for a car to persuade its driver to become more eco-friendly? Can new technology in cars actually prevent drivers from succumbing to road rage?
Given the rapid frequency with which social science and technology collide, Nass hasn’t had any difficulty finding research questions for his course. While it may seem like a tall order, many of the students in the class do find research solutions and even have the published studies to show for it.
The class works in groups to address questions about technology, allowing students to approach problems from multiple disciplinary perspectives.
“Everyone comes from different backgrounds and that variety makes solutions possible,” Song said.
Nass wanted groups to span the “techie-fuzzy” divide. Working separately is “just not real life…in any job virtually everyone works in teams,” he said.
Since most of the class’ research questions are fairly straightforward, students who have no experience in computer science can do meaningful research and learn as they go. In other words, the goal of the course, as Nass put it, is to “not let their ignorance be a barrier.” Unlike pure academic disciplines like math and physics, which have been more uniformly researched, changing technology means many simple, new, unanswered questions are waiting to be explored.
The combination of group work and opportunity makes discovery so close students can almost taste it, and many do. Students with successful studies from in the class often have their papers published, sometimes before they complete their undergraduate degrees. Basically, students “don’t have to be experts to make huge contributions,” according to Nass.
“Comm 168…makes research accessible for a huge number of students,” Dole said.
A typical example of how research in the class is conducted is the “Amanda” experiment. Their group meeting, which was typical of the class’s format, consisted of brainstorming possible problems the study might encounter and finding solutions to address them. In the case of “Amanda,” the pitch variants of male and female voices were very problematic.
Most the meeting consisted of listening to synthetic voices and playing the statements at warp speed and slower speeds on the computer, analyzing the different factors that pace and volume play in speech.
The famous Stanford entrepreneurial spirit feels present in Nass’s course, where two heads may be better than one.