A new study published in PLOS ONE journal analysed a corpus made up of 700 million words and phrases which appeared on the Facebook accounts of 75,000 volunteers who participated in the research. Specialists used an open vocabulary technique seeking to eliminate the limitations associated with closed-vocabulary analyses, turning it into the largest study in its kind. The study was carried out by researchers from University of Pennsylvania’s Positive Psychology Center and it was directed by H. Andrew Schwartz, PhD in computer and information science. The investigation is a part of the World Well-Being Project, which is testing new techniques for measuring well-being through social media.
Limitations of a Closed Vocabulary Approach
Throughout the years, researchers have studied the relationship between people, language and the way they comprehend feelings and reality. However, the approach chosen in these studies is generally a closed approach, which severely limits the results of the study, stated Margaret Kern, one of the specialists involved in the study. Instead of analysing the entire linguistic corpus, studies on similar topics to this one tend to choose a list of relevant words and then investigate the frequency with which they are used. An important problem related to this method is that the selection of words is limited or influenced by the results that the researchers are expecting to find or by the established preconceptions.
Open Vocabulary Approach and Conclusions
This new study by the Positive Psychology Center at Pennsylvania has chosen to derive keywords from the sample itself. Words and phrases were isolated in clusters, each related to specific characteristics such as gender, age or personality traits (extraversion, agreeableness, conscientiousness, neuroticism and openness). The information was translated visually into word clouds which easily display the words and their correlation to the cluster represented.
The importance of the study lies on its potential to derive new insights into the relationship between language and traits. Lyle Ungar, another professor involved in the research, has stated that some of the results show a correlation between words related to active and social activities, and people with greater emotional stability. According to her, this could mean various different things, but the main issue is the implied suggestion that the topic should be investigated further.
The World Well-Being Project
The World Well-Being Project involves a group of computer scientists and statisticians who cooperate with psychologists to investigate psychological processes, seeking to create new, cheaper, and more efficient measurements of well-being and other related methods of information seeking. The project’s final aim is to promote policy making which is up to date and responsive to society’s interests and well-being.