MINING ERROR VARIANCE AND HITTING PAY-DIRT - DISCOVERING SYSTEMATIC VARIATION IN SOCIAL SENTIMENTS

According to traditional error theory, sentiment measurements vary uns ystematically from individual to individual. However, we find some pat terned deviation in sentiments that characterize subsets of respondent s within a seemingly homogeneous population. After demonstrating-the e xistence of such patterns, we report an exploratory study aimed at ide ntifying social characteristics of people with different patterns of s entiments. People embedded in multiple social networks have exaggerate d sentiments that contour cultural patterns, and people with few socia l associations have attenuated sentiments conveying cultural detachmen t. In addition, people with disciplinary parents and romantic privatio n have some distinctive sentiments, depending on their gender.