SELF-CONTROL IN A CRIMINAL SAMPLE - AN EXAMINATION OF CONSTRUCT-VALIDITY

Favorable evidence on the validity of the Grasmick et al. (1993) self- control scale has been reported in studies using general population sa mples. However, the scale has never been tested among persons extensiv ely involved in crime. We assessed the construct validity of this scal e, slightly revised, in a heterogeneous sample of drug-using criminal offenders. Factor analyses identified five subscales, mostly congruent with existing formulations of the self-control construct. Also, recen t crimes of force and fraud were more frequent among people scoring lo wer on self-control. However, the five-factor solution was not tenable among women, and the scale was no more closely related to crime than were three subscales representing more specific constructs already est ablished in criminology.