Self-Control Assessments and Implications for Predicting Adolescent Offending

Adam Fine, Laurence Steinberg, Paul J. Frick, Elizabeth Cauffman

Research output: Contribution to journalArticlepeer-review

30 Scopus citations


Although low self-control is consistently related to adolescent offending, it is unknown whether self-report measures or laboratory behavior tasks yield better predictive utility, or if a combination yields incremental predictive power. This is particularly important because developmental theory indicates that self-control is related to adolescent offending and, consequently, risk assessments rely on self-control measures. The present study (a) examines relationships between self-reported self-control on the Weinberger Adjustment Inventory with Go/No-Go response inhibition, and (b) compares the predictive utility of both assessment strategies for short- and long-term adolescent reoffending. It uses longitudinal data from the Crossroads Study of male, first-time adolescent offenders ages 13–17 (N = 930; 46 % Hispanic/Latino, 37 % Black/African-American, 15 % non-Hispanic White, 2 % other race). The results of the study indicate that the measures are largely unrelated, and that the self-report measure is a better indicator of both short- and long-term reoffending. The laboratory task measure does not add value to what is already predicted by the self-report measure. Implications for assessing self-control during adolescence and consequences of assessment strategy are discussed.

Original languageEnglish (US)
Pages (from-to)701-712
Number of pages12
JournalJournal of youth and adolescence
Issue number4
StatePublished - Apr 1 2016
Externally publishedYes


  • Adolescent delinquency
  • Impulsivity
  • Self-control

ASJC Scopus subject areas

  • Social Psychology
  • Education
  • Developmental and Educational Psychology
  • Social Sciences (miscellaneous)


Dive into the research topics of 'Self-Control Assessments and Implications for Predicting Adolescent Offending'. Together they form a unique fingerprint.

Cite this