![[ Visit Client Website ]](images/banner.JPG)
Thursday, May 7, 2009
Northwest Hall (Chicago Hilton)
2:30 PM
Background:
Significant work has focused on valid and reliable identification of autistic spectrum disorders (ASDs), as well as its evidence-based treatment. However, there remains a gap in moving systematically from signs of ASDs to individualized selection of treatments based on these clinical characteristics. We report on development of the ALD-I, a measure designed to operationalize autistic symptomatology so treatments can be selected that target a S’s specific key deficits. To accomplish this, ASDs were reconceptualized as a constellation of autism-specific learning disabilities (ALD) and autism-specific learning styles (ALS) that can then be mapped onto treatments that target deficits in key domains. Doing this can facilitate outcome research where groups can be matched or co-varied according to ALD-I profile, or outcome can be examined with respect to specific ALD.
Objectives:
We will present factor analytic data supporting the validity of the ALD-I for identification of key domains of ASDs. Cohesive clinical sub-domains on each factor will be described with respect to independence, internal consistency, and variance accounted. Procedures in instrument development (e.g., item inclusion, deletion and scaling) will be reported.
Methods:
The Autistic Learning Disabilities Inventory (ALD-I) (N=164) was completed by primary caregivers of Ss age 4 – 17 years referred for ASD evaluation. The initial ALD-I was formulated as a 68-item questionnaire using a 5-point Likert Scale. The items were theoretically clustered into 3 domains: Social, Communication and World of Objects. Within each domain 3-4 sub-domains scaled the items into specific skill areas (e.g., Imitation). For item selection/ retention analyses, Ss were grouped into two diagnostic categories: Autism (AD) and ASD (Autism, Asperger's, and PDD-NOS). Comparative data were also collected on ASD-referred Ss later found to meet criteria for mental retardation, ADHD or a language disorder. A further 225 comparable cases received ALD-Is with simplified 3-point scaling analyzed later.
Results:
The goals of data analysis were to 1) derive factors to identify empirical ALD and ALS domains and sub-domains, 2) describe internal consistency of sub-domains, 3) calculate differential (mean, standard deviation) scores on these factors and sub-domains in study and comparison groups, and 4) eliminate redundant, unreliable and non-loaded items. Initial rotated factor analysis of 5 point data identified 3 distinct factors with factor loadings ≥ .35- I: 'World of Objects,' II: 'Social and Non-Verbal Communication' and III: 'Verbal' (with respective variance accounted= 7.5, 7.2, and 4.9). A comparison of factor score sums across diagnostic groups revealed significant differences between AD, ASD and comparisons, as did parallel analyses of sub-domains for each factor.
Conclusions:
The ALD-I measures a coherent set of social, verbal and play skills useful in identifying specific learning deficiencies as well as relative strengths for ASD children. The measure should be useful in 1) identifying Ss with deficits likely to respond to specific treatments, 2) identifying relative strengths that can be capitalized upon for treatment, and 3) designing treatment research where explication of characteristics of responders is valuable.
Significant work has focused on valid and reliable identification of autistic spectrum disorders (ASDs), as well as its evidence-based treatment. However, there remains a gap in moving systematically from signs of ASDs to individualized selection of treatments based on these clinical characteristics. We report on development of the ALD-I, a measure designed to operationalize autistic symptomatology so treatments can be selected that target a S’s specific key deficits. To accomplish this, ASDs were reconceptualized as a constellation of autism-specific learning disabilities (ALD) and autism-specific learning styles (ALS) that can then be mapped onto treatments that target deficits in key domains. Doing this can facilitate outcome research where groups can be matched or co-varied according to ALD-I profile, or outcome can be examined with respect to specific ALD.
Objectives:
We will present factor analytic data supporting the validity of the ALD-I for identification of key domains of ASDs. Cohesive clinical sub-domains on each factor will be described with respect to independence, internal consistency, and variance accounted. Procedures in instrument development (e.g., item inclusion, deletion and scaling) will be reported.
Methods:
The Autistic Learning Disabilities Inventory (ALD-I) (N=164) was completed by primary caregivers of Ss age 4 – 17 years referred for ASD evaluation. The initial ALD-I was formulated as a 68-item questionnaire using a 5-point Likert Scale. The items were theoretically clustered into 3 domains: Social, Communication and World of Objects. Within each domain 3-4 sub-domains scaled the items into specific skill areas (e.g., Imitation). For item selection/ retention analyses, Ss were grouped into two diagnostic categories: Autism (AD) and ASD (Autism, Asperger's, and PDD-NOS). Comparative data were also collected on ASD-referred Ss later found to meet criteria for mental retardation, ADHD or a language disorder. A further 225 comparable cases received ALD-Is with simplified 3-point scaling analyzed later.
Results:
The goals of data analysis were to 1) derive factors to identify empirical ALD and ALS domains and sub-domains, 2) describe internal consistency of sub-domains, 3) calculate differential (mean, standard deviation) scores on these factors and sub-domains in study and comparison groups, and 4) eliminate redundant, unreliable and non-loaded items. Initial rotated factor analysis of 5 point data identified 3 distinct factors with factor loadings ≥ .35- I: 'World of Objects,' II: 'Social and Non-Verbal Communication' and III: 'Verbal' (with respective variance accounted= 7.5, 7.2, and 4.9). A comparison of factor score sums across diagnostic groups revealed significant differences between AD, ASD and comparisons, as did parallel analyses of sub-domains for each factor.
Conclusions:
The ALD-I measures a coherent set of social, verbal and play skills useful in identifying specific learning deficiencies as well as relative strengths for ASD children. The measure should be useful in 1) identifying Ss with deficits likely to respond to specific treatments, 2) identifying relative strengths that can be capitalized upon for treatment, and 3) designing treatment research where explication of characteristics of responders is valuable.