Psychometric properties of a short self-report measure of rule-breaking behaviour among adolescents: findings from the Ungdata survey

BACKGROUND The aim of the present study was to examine the factor structure and reliability of a six-item scale of rule-breaking behaviour, and to test for measurement invariance across gender, age, survey year and geographical location. METHODS Data were from three yearly cross-sectional and population-based collections of the Ungdata surveys (2017 to 2019) including a total of 297,102 Norwegian adolescents aged approximately 13 to 19 years. Measurements included respondent's rule-breaking behaviour, time, gender, age and geographical location. RESULTS Confirmatory factor analyses demonstrated that a one-factor solution of the rule-breaking behaviour scale had good fit to data (comparative fit index 0.98; Tucker-Lewis index 0.96; root mean square error of approximation 0.049 (95% confidence interval 0.048, 0.050)), with factor loadings ranging from 0.60 to 0.81 for all items (mean factor loading 0.72). Similar results were found across survey years for both genders. Several multiple group confirmatory factor analyses showed indications of measurement invariance for the scale across gender, age groups, geographical locations and survey years. The ordinal alpha and omega coefficients for internal consistency of the scale were both 0.86. CONCLUSIONS The six-item scale for self-reported rule-breaking behaviour demonstrated good psychometric properties and appears to constitute a reliable measure of adolescent rule-breaking behaviour for use in population-based surveys in a Norwegian setting.