Research in Autism Education

Category Description

Submissions for acceptance in this category will be evaluated based on scientific merit, quality of design and control, and relevance to the fields of autism assessment, intervention, staff training, basic conceptual understanding, and program design. Additional consideration will be given based on the relevance of the research topic to the scientific literature/research base in the fields of special education, autism treatments, psychology, and behavior analysis. Consideration will be given to the degree that experimental control is demonstrated, including threats to internal and external validity as well as reliability of dependent variable measurement. Preference will be given to studies that include measures of inter-observer agreement for the dependent and independent variable, as well as data related to treatment fidelity.

Submission Criteria

Experimental/Quasi-experimental design including single subject designs and inter-group designs. These proposals will demonstrate systematic manipulations and test functional relations either between groups, across individuals, or within an individual.

1. The proposal will include definitions of both dependent and independent variables.
2. Various single subject designs used in isolation or in various combinations can be used and must be explicitly stated (i.e., reversal designs, multiple baseline, alternating treatments, and changing criteria designs).
3. For group designs, the proposal will need to include evidence of an inter-group experimental design to test functional relations and must include clear hypothesis statements, a description of population sample, and a clear statement of sample size and methods of analysis.
4. Baseline measurements of dependent variables are required for all submissions.
5. Preference will be given to proposals that include IOA and treatment integrity data.
6. If study involves problem behavior reduction, Functional analysis, or at a minimum FBA data, is required.
7. This category does not include data derived primarily from verbal reports. Direct measures must be included for consideration.
8. Data will be included and presented in a form that allows visual analysis. Posters should include graphs and charts allowing a fairly efficient analysis of the data. Inter-group design studies are to include clear charts summarizing experimental data including any statistical analysis.

Sample Proposal

Title of Poster Presentation: Comparing use of prompting strategies to teach picture labeling for two children with autism

Primary Presenter Contact: Katelyn Butler

Role: Special Education Consultant

Additional Presenter: Isabelle Crawford

Role: Special Education Teacher

School District/Intermediate Unit/Agency: Western School District

Poster Description/Topic: This study was completed to compare the effects of two prompting techniques, full echoic and partial echoic, on the acquisition of vocal picture labeling for two children with autism. The study is a partial replication of work completed by Parr and LeBlack, 2011.

Subjects: Two school age students with Autism. Both students performed at Level 1 for picture labeling on the Competent Verbal Scale for Students with Autism, Elementary Level. Both students are vocal learners and have developed the ability to repeat what has been said to them. Informed consent from the student’s families was obtained and all personally identifying information for the children and their families has been withheld.

Study Design: The study includes an alternating treatment single subject design, replicated across the two students. Baseline data was collected on vocal labeling skills for three sets of three picture stimuli presented in a discrete trial format. For the first set each child was taught each item using a full echoic prompt. The second set was taught using a partial echoic prompt following one trial with a full echoic model of the response. For both conditions, each prompted trial was followed by an unprompted probe trial to assess acquisition. Each set was taught one time per day for 30 trials. The order of teaching was randomly varied throughout the study. The dependent variable, number of days to mastery for each item in the set, was defined through trial by trial data resulting in two consecutive days of 90% correct or better for each item in each set on unprompted trials. The independent variable for each condition was instruction through discrete trial instruction procedures and varying type of prompt presented by treatment condition.

Measurement of Independent and Dependent Variable: For dependent variable two independent observers scored performance on probe trials for 25% of all baseline and treatment sessions over the course of the study. For the independent variable, the instructor providing training completed a treatment integrity checklist following each session for all sessions. An independent observer completed the checklist based on instructor behavior for 25% of all treatment sessions.

Data Display: Data will be presented through an equal interval graph that includes baseline and treatment phase conditions. Separate data paths for each treatment condition will be shown for each student for treatment phase. Here is preliminary data for one subject: