Using Item Response Theory Methods With the Brazilian Temperament Scale for Students

Abstract

The development and validation of the Brazilian Temperament Scale for Students (BTSS) are examined through the use of data from 1,258 children and adolescents, ages 10 through 21 ( M = 15.0, SD = 2.1, 56% females). Three psychometric properties of BTSS are reported: its internal structure (e.g., validity), its reliability, and cut points to best distinguish between the bipolar styles for each of the scale’s four constructs (extroversion–introversion, practical–imaginative, thinking–feeling, and organized–flexible styles). Rasch analyses were used to estimate item intensity and students latent score parameters. The use of construct maps to help establish norms is described. The results of exploratory factor analysis using varimax rotation confirm the four scales’ intended bipolar factors are composed of items that theoretically represent the desired constructs. The confirmatory factor analysis for a four-factor structure generally displays good fit indexes. Infit and outfit values reach acceptable ranges (i.e., from 0.68 to 1.32). Results are interpreted using cut points.