Self-control in children: a multimethod examination of treatment outcome measures.

This investigation consisted of a multimethod evaluation of treatment outcome measures that have been developed to assess self-control in children. Subjects were 132 children from the fourth, fifth, and sixth grades. Each child's homeroom teacher and one other classroom teacher independently completed the Self-Control Rating Scale (SCRS), the Teacher's Self-Control Rating Scale (TSCRS), and the 10-item Conners Teacher Rating Scale (TRS). Parents of 41 children completed the Teacher's Self-Control Rating Scale and the Conners Teacher Rating Scale. Children completed the Children's Perceived Self-Control Scale (CPSC) and were administered the Matching Familiar Figure Test (MFF). Results showed high internal consistency reliability for the TSCRS, SCRS, TRS, and MFF latency scores. CPSCS and MFF error scores demonstrated low reliability (r alpha's = .58 and .61, respectively). Interrater reliability between teachers was considered acceptable for the TSCRS but not for the SCRS and TRS. Correlations between teacher and parent ratings on the TSCRS and TRS were low (r's = .37 to .50). Correlations between children's latency scores on the MFF and teachers and parent ratings were very low, r's = .01 to .19, as was the correlation with children's self-ratings on the CPSCS (r = .07). Correlations between rating scales and MFF error scores also were low. The findings suggest that considerable construct confusion does exist between measures designed to assess self-control and impulsivity in children.