| Literature DB >> 25272448 |
Hiroyuki Ito, Kaori Matsumoto, Nobuya Takayanagi, Shin Harada, Satoko Ohtake, Naoto Mochizuki, Syunji Nakajima, Wataru Noda, Yoshihiro Tanaka, Masatsugu Tsujii.
Abstract
We examined the psychometric properties of the Japanese version of the Autism Spectrum Screening Questionnaire (ASSQ) and developed a short-form. This study included 157 children with autism spectrum disorders (ASD, ages 7-18, 128 boys) and 4,101 healthy controls (ages 7-15, 3,344 boys) from a general population with a controlled male-female ratio. Four factors (Unusual Interests, Sociality, Peer Relations, and Repetitive Behaviors) were extracted by exploratory factor analysis of control group data. Confirmatory factor analysis revealed that the 4-factor model fit well with data for another sample of the control and ASD groups. Logistic analysis showed that the former 3 factors could significantly predict ASD diagnosis. Thus, a short form of the ASSQ was developed, consisting of 11 items for these 3 factors. This short form showed sufficient internal consistency and high discrimination power for ASD diagnosis that was comparable to that of the 22-item version. Receiver operating characteristic analysis indicated an optimal cut-off of 7 for the 22-item version (sensitivity .949, specificity .801) and 5 for the short-form (sensitivity .936, specificity .818).Entities:
Mesh:
Year: 2014 PMID: 25272448 DOI: 10.4992/jjpsy.85.13213
Source DB: PubMed Journal: Shinrigaku Kenkyu ISSN: 0021-5236