Reliability of Child Diagnosis in a Hispanic Sample

Abstract This study presents data on test-retest diagnostic reliability between two child psychiatrists. Independent ratings were obtained on both community and clinical subjects on a diagnostically heterogeneous sample of Puerto Rican children. Reliability was better at the higher rank of major categories than at the level of specific diagnoses. Age or time lapse between the interviews did not affect diagnostic reliability. Agreement was significantly better on either the referred severe cases or the nonreferred, noncases. Issues related to the taxonomy of childhood disorders are discussed.