Measurement of behaviour difficulties in children using standard scales administered to mothers by computer: Reliability and validity.
Measurement of behaviour difficulties in children using standard scales administered to mothers by computer: Reliability and validity.
复制标题
DOI：
10.1007/bf02084430
复制
发表时间：
1992-01-01
影响因子：
6.4
通讯作者：
McGuire, Ralph
中科院分区：
文献类型：
--
作者：
Berg, Ian;Lucas, Christopher;McGuire, Ralph
72 child psychiatric and 50 paediatric outpatients were included in a study looking at the Rutter A (2) Scale and the Child Behaviour Checklist (CBCL) completed by mothers. Automated administration with recorded voice as well as visual display of written questions was compared to standard administration: pen and paper with the Rutter Scale and a psychologist asking the CBCL questions. A proportion of each group of cases was re-assessed some weeks later. Each sample of cases was randomly assigned to one of four groups (1st occasion-2nd occasion): computer-computer, standard-standard, computer-standard and standard-computer. Within each, order of presentation of the scales was also randomly allocated. Analyses of variance did not reveal differences in Rutter or CBCL total scores between the computer and standard methods of administration at the time of the initial assessment. Intraclass correlation coefficients indicated high test-retest reliabilities within the 4 subgroups. ROC (Relative Operating Characteristics) analysis showed both scales to be effective in discriminating between paediatric and psychiatric cases with cut-off points of 13 or more on the Rutter Scale and 42 or more on the CBCL, using either computer or standard methods of administration. For routine clinical use the Rutter A (2) Scale administered by computer is to be preferred because of its good test-retest reliability, high concurrent and criterion validity, as well as its brevity, acceptability and ease of administration.
