Introduction
Past research has demonstrated that children with Autism and Autism Spectrum Disorders benefit from early intervention utilizing Applied Behavior Analysis techniques. Much of this research has primarily focused on children 3-4 years of age receiving between 10-40+ hours per week of intervention (Kasari, 2002). Thus far, research has not conclusively indicated how much early intervention is needed, or at what age services should begin. Outcome data were analyzed from an in-home early intervention program for children under 3 with Autism or PDD-NOS. Children entered the program at various points of time, received 12 hours per week of 1:1 intervention, and exited at age 3. Outcome data at program end representing cognitive, language, and social/emotional measures from the Bayley and the Brigance were compared across children grouped by program duration. Bayley data were analyzed controlling for mental retardation. Results were compared to children of the same age who had received no prior intervention.

Methods
Subjects: Children were selected for analysis based on age, testing data, and duration of program. “Control” children entered the program at 34-35 months of age and had no prior intervention. Children were separated into groups based on program duration
Bayley Participants: Two analyses were conducted on Bayley data: High MDI and Low MDI. The cut-off of MDI score ≥ 65 was used for the High MDI group to control for possible effects of mental retardation. The Low MDI group comprised all children in the program with MDI scores to < 50. Children were separated into groups based on program duration of 4 months.
Brigance Participants: 372 children comprised the Brigance group. Children were separated into groups based on program duration of 3 months. Language composite scores were created from the Brigance as follows:
- Receptive Composite: Receptive Colors, Receptive Body Parts, Receptive Vocabulary
- Expressive Composite: Expressive Colors, Expressive Body Parts, Expressive Vocabulary

Early Intervention Program: Children received 12 hours/week of home-based 1:1 intervention. The program implements both Discrete Trial and Pivotal Response Teaching techniques. Interventionists have either their BA or MA.

Analyses: Scores from the Bayley were analyzed using univariate ANOVAs with program duration as the between-subjects factor and MDI as the within-subject factor. Pairwise comparisons, corrected for multiple comparisons, were utilized to assess group differences. Means of Brigance data were used and represented graphically.

- Post-test data was used for children who had received intervention for 2 or more months; pre-test data was used for children who entered the program at 34 months of age or later.

Results-Bayley
High MDI: At a mean age of 35 months, there was a significant effect of program duration on MDI score (F (3, 73) = 11.685, p = .000 ).
- Post-hoc tests revealed that children enrolled in the program 2-6, 6.1-10, and 10.1-14 months scored significantly higher compared to the control group (Figure 1 and Table 1).
- Children receiving intervention for 10.1-14 months scored significantly higher than children who received 2-6 and 6.1 to 10 months of services.
Low MDI: At a mean age of 35 months, there was a significant effect of program duration on MDI score (F (3, 80) = 5.470, p = .002 ).
- Post-hoc tests revealed that children enrolled in the program 2-6, 6.1-10, and 10.1-14 months scored significantly higher compared to the control group (Figure 2 and Table 1).

Results- Brigance
At a mean age of 35 months, age equivalent scores for children receiving intervention were greater than those of controls.
Language: Both Receptive and Expressive Composite scores were greater for children receiving intervention than for controls. Gains increased as a function of program duration. Children receiving intervention exhibited greater expressive than receptive composite scores, whereas controls exhibited the reverse trend (Figure 3).
Social/Emotional: Social/Emotional scores were greater for children receiving intervention than for controls. Gains increased as a function of program duration (Figure 4).

Conclusion
- Children with MDIs ≥ 65 receiving intervention scored significantly higher on the Bayley compared to a control group; the longer a child participated in the program the greater the gains.
- These effects remained after children with lower MDIs were included, suggesting that the significant differences are a true effect of intervention and are not accounted for by cognitive level.
- Children receiving intervention demonstrated greater language and social/emotional skills than controls at 35 months of age. Expressive skills appears to be the area of greatest gain.
- These results indicate that early intervention assists in the acquisition of skills and longer periods of intervention are associated with greater gains. Additional outcome studies are necessary to assess the effects of early intervention programs for young children with Autism and what child-attributes may contribute to gains made.