The stereotyped behavior scale: psychometric properties and norms

Citation
J. Rojahn et al., The stereotyped behavior scale: psychometric properties and norms, RES DEV DIS, 21(6), 2000, pp. 437-454
Citations number
34
Categorie Soggetti
Rehabilitation
Journal title
RESEARCH IN DEVELOPMENTAL DISABILITIES
ISSN journal
08914222 → ACNP
Volume
21
Issue
6
Year of publication
2000
Pages
437 - 454
Database
ISI
SICI code
0891-4222(200011/12)21:6<437:TSBSPP>2.0.ZU;2-Q
Abstract
The Stereotyped Behavior Scale (SBS) is an empirically developed behavior r ating scale for adolescents and adults with mental retardation (Rojahn, Tas se & Sturmey, 1997). Since the original publication, one item was deleted a nd two items were merged, leaving 24 items. In an additional change, severi ty scales were added to the frequency scales. In this paper, psychometric p roperties and (relative) norms for the new SBS are presented. In the psycho metric study, 45 adults with mental retardation from a residential facility participated. Of these, 15 were selected for high-rates or very severe for ms of stereotyped behaviors, 15 for mild to moderate rates or less severe s tereotypies, and 15 for the low rates or absence of stereotyped behaviors. Direct care staff familiar with the participants completed the SBS and the "Stereotypy" subscale of the Aberrant Behavior Checklist-Residential (ABC-R ) (Aman, Singh, Stewart & Field, 1985). For 15 participants, two raters ind ependently completed the SBS. In addition, 45-min direct behavior observati ons were conducted on 16 participants. After approximately four weeks: the instruments were completed a second time by the same raters. As for reliabi lity, the SBS frequency and severity scale total scores yielded test-retest intraclass coefficients (ICC) of 0.93 and 0.71, ICC interrater agreement o f 0.76 and 0.75, and each had an internal consistency alpha of 0.91. For cr iterion validity, the SBS frequency and severity scores correlated with the ABC-R "Stereotypy" score at 0.80 and 0.84 (Pearson r), with systematic beh avior observations at 0.50 and 0.65 (Pearson r), and with the a priori clas sification at 0.50 and 0.65 (Spearman rho). From a previous data set of 550 individuals with stereotypic behavior, normative data (percentile ranks an d T-scores) were derived. The data were presented in two tables, one showin g a breakdown of gender by age groups, and the second one of age groups by level of functioning. (C) 2000 Elsevier Science Ltd. All rights reserved.