![[ Visit Client Website ]](images/banner.JPG)
Thursday, May 15, 2008
Champagne Terrace/Bordeaux (Novotel London West)
11:30 AM
Background: The Autism Diagnostic Interview–Revised (ADI-R) is one of the most widely used assessment instruments in the field. Relatively few studies have used factor analysis to assess its items. Such analyses could shed some light on the structure of autistic symptoms and on the validity of the diagnostic algorithm.
.
Objectives: The purpose of the current study was to investigate further the factor structure of the ADI-R. Exploratory (EFA) and confirmatory factor analysis (CFA) were conducted on the ADI-R items (algorithm items only and all items).
Methods: Data were obtained from the Autism Genetic Resource Exchange program. The sample included 1,861 youngsters with PDDs between the ages of four and 18 years (mean=8.3, SD=3.2). The sample consisted of 1,455 males (78%) and 406 females (22%). Analyses were conducted separately for algorithm items only and for all items and according to verbal status (n=1329 verbal and n=532 non-verbal). For all analyses, EFA (Ordinary Least Squares and oblique rotations) was conducted first on a random subsample and then followed by CFA on the entire sample. Several models were compared with the CFAs.
Results: Overall, results indicated a two-factor solution best fit all four data sets (algorithm items only/all items, verbal/nonverbal children). The first factor consisted of social items and the majority of communication items. The atypical communication items (e.g., stereotyped utterances, pronominal reversal, neologisms) and repetitive behaviors loaded onto the second factor. CFAs suggested excellent fit indices for most solutions (i.e., RMSEA<.05; SRMR<.09).
Conclusions: Data suggested that autism symptomatology can be explained statistically with a two-domain model. Analyses permitted the identification of which symptoms are most/least correlated with different domains. Factor analytical studies of this nature can assist in refining the autism phenotype and improving the diagnostic algorithm.
.
Objectives: The purpose of the current study was to investigate further the factor structure of the ADI-R. Exploratory (EFA) and confirmatory factor analysis (CFA) were conducted on the ADI-R items (algorithm items only and all items).
Methods: Data were obtained from the Autism Genetic Resource Exchange program. The sample included 1,861 youngsters with PDDs between the ages of four and 18 years (mean=8.3, SD=3.2). The sample consisted of 1,455 males (78%) and 406 females (22%). Analyses were conducted separately for algorithm items only and for all items and according to verbal status (n=1329 verbal and n=532 non-verbal). For all analyses, EFA (Ordinary Least Squares and oblique rotations) was conducted first on a random subsample and then followed by CFA on the entire sample. Several models were compared with the CFAs.
Results: Overall, results indicated a two-factor solution best fit all four data sets (algorithm items only/all items, verbal/nonverbal children). The first factor consisted of social items and the majority of communication items. The atypical communication items (e.g., stereotyped utterances, pronominal reversal, neologisms) and repetitive behaviors loaded onto the second factor. CFAs suggested excellent fit indices for most solutions (i.e., RMSEA<.05; SRMR<.09).
Conclusions: Data suggested that autism symptomatology can be explained statistically with a two-domain model. Analyses permitted the identification of which symptoms are most/least correlated with different domains. Factor analytical studies of this nature can assist in refining the autism phenotype and improving the diagnostic algorithm.