The Children's PTSD Inventory: development and reliability

Abstract

Information involving the development of the DSM-IV version of the Children's PTSD Inventory is described. Independent ratings by highly experienced judges denote that the instrument encompassed the universe of definition that it was intended to measure (i.e., the DSM-IV criteria for PTSD). The instrument was administered to 82 traumatized and 22 nontraumatized youths at Bellevue Hospital. Moderate to high Cronbach alphas (.53-.89) were evident at the subtest level. An alpha of .95 was evident at the diagnostic level. In terms of inter-rater reliability, 98.1% agreement was evident at the diagnostic level. Inter-rater intraclass correlation coefficients (ICCs) ranged from .88 to .96 at the subtest level and .98 at the diagnostic level. Good to excellent kappas (.66-1.00) were reported for inter-rater reliability at the subtest level. An inter-rater reliability kappa of .96 was evident at the diagnostic level. In terms of test-retest reliability, 97.6% agreement was evident at the diagnostic level. Good to excellent test-retest kappas (.66-1.00) and ICCs (.66-.94) were observed. A test-retest kappa of .91 and an ICC of .88 was observed at the diagnostic level.