Annotation and personality: Individual differences in sentence boundary detection

Abstract

The paper investigates the relationships between personality traits and expert manual annotation of unscripted speech. The analysis is based on the results of a psycholinguistic experiment in which the participants were asked to annotate sentence boundaries in transcriptions of Russian spoken monologues. Personality traits were measured using the Eysenck Personality Inventory and the Five Factor Personality Questionnaire. A multiple regression model showed that the traits measured by the Five Factor Personality Questionnaire along the scales ‘unemotionality vs. emotionality’ and ‘practicality vs. playfulness’ had significant effect on the average sentence length.