Lisa Feldman Barrett is looking for happiness (and sadness)

Some 300 Boston College students have been subjects in an experiment that is likely to influence not only the way psychologists understand human emotions but also the way they go about studying them. Much of the data from the research, conducted by Associate Professor Lisa Feldman Barrett, comes from the students who carry Palmtop computers everywhere they go. When the Palmtops beep--which usually happens about 10 times a day--the first question that appears on the screen is "Are you having an emotion?" This is followed by a random series of queries, starting with, say, "How happy are you right now?" and inquiring into any of 28 additional emotions. The students have agreed to be on call to answer questions about their emotional states 15 hours a day, for four weeks at a time. The project is supported by a $300,000 grant from the National Science Foundation.

Psychologists historically have had a complicated relationship with emotion studies. Though the first psychology textbook--William James's The Principles of Psychology (1890)--contained a chapter on emotion, early psychologists preferred to focus on behavior. There were few studies of emotion before the 1970s and no strong surge of interest until the 1990s. The challenge has been one of measurement: how do you take an individual's subjective state and present it in a way that permits comparisons? Barrett has devised ways to do so that are new to the discipline.

"I'm not really interested at any given instant in whether someone is happy or sad," says Barrett of her research. "What I'm interested to know is, if they're happy, do they distinguish between, say, happy, enthusiastic, and calm--or do they lump it all together as feeling good?" The term she has come up with for an individual's ability to make fine distinctions among emotions is "granularity." ("In psychology," she chuckles, "you have to have cool terms.") Someone who reports feeling sad whenever feeling angry will have a lower granularity rating than a person whose anger is accompanied sometimes by sadness and other times by, say, frustration.

Psychologists have traditionally thought of emotions as discrete phenomena (anger is distinct from sadness, which is distinct from fear, for instance). Barrett posits that emotions start as elementary good or bad feelings, which become refined by what a person knows from experience, by states of physical arousal, and by cognitive capacity. In her experiment, she uses the Palmtop data to assign a granularity rating to each subject. Then, in their weekly visits to the lab, she asks subjects to perform tasks that will help her to determine whether and to what extent granularity ratings correlate with other characteristics.

The key to the success of the study is the Palmtop. "If you ask people, 'Hey, can you tell the difference between your emotional states?'" says Barrett, "the answer you're likely to get is a theory rather than what they actually do." The Palmtops' questions, which require a response within two minutes, provide data that is reflexive, subtle, and more likely valid. According to Yale University's Peter Salovey, researchers have been doing so-called experience sampling, relying on paper and pen reports by subjects outside the laboratory, for about 15 years. Barrett, he says, "is one of the first to apply the approach to emotions, and certainly the first to do it on a Palmtop. In the field of emotions today, there may be no more sophisticated methodologist and measurement expert." Barrett plans to submit the first results of her work for publication this summer.

The software that makes Barrett's study possible was developed by her husband, Daniel Barrett, a computer scientist. Aided by a supplementary grant from the National Science Foundation, the couple has made their Experience Sampling Program, called ESP, available at Barrett's Web site http://www2.bc.edu/~barretli/esp for free. "Experience sampling is very time-consuming and very expensive," says Barrett. "I would go to conferences and realize that people were having a lot of trouble with it--you know, not everyone has a computer programmer at their disposal."

Professor Salovey describes the Barretts' gesture as "wonderfully generous. Often when people develop something like this they prefer to sell it." ESP, he says, will make it easier for other emotion researchers "to be at the cutting edge."

Anna Marie Murphy

Photos: Barett: Portable computers give psychologists human emotions data they can trust. Gary Wayne Gilbert


Top of page