Emergent Multi-Class Imitation Training (EMIT)

The goals of this project are to (a) incorporate empirical findings on imitation training and learning in autistic children into a comprehensive protocol for Applied Behavior Analysis practitioners designed to rapidly facilitate generalized imitation called Emergent Multi-Class Imitation Training (EMIT), and (b) collect pilot data on the efficacy of EMIT with a small sample of autistic children. EMIT will incorporate several features that are grounded in prior research including: (a) evidence-based procedures for establishing trained matching relations (a pre-requisite to generalized imitation), (b) concurrent training of different response types (e.g., motor imitation, object imitation, vocal imitation) to address restricted generalization, (c) multiple manipulative object imitation training, (d) evidence-based procedures for remediating slow acquisition, and (e) frequent tests for the emergence of generalized imitation. EMIT will be the first protocol designed for clinical use that reflects research findings on imitation learning spanning almost five decades.

Using findings from prior research, the investigators will design a manualized training protocol (EMIT) for establishing generalized imitation in autistic children that supervised behavior therapists can implement with a high degree of fidelity. The EMIT protocol will serve as a guide for behavior therapists to help them deliver high-quality, empirically derived imitation training to autistic children. The investigators will conduct a small pilot study to evaluate the effects of EMIT on the emergence of generalized imitation with young autistic children. Single-case experimental design and a 2x2 group design will be used. A multiple baseline across participants design is a single-case design that is common in imitation research with this population. In short, the start of the intervention is stagged across participants to systematically assess the effects of the intervention on each participant's behavior. In this experiment, confirmation that EMIT is responsible for any observed effects will be obtained when changes in each participant's behavior (i.e., correct imitation of novel responses) occur when, and only when, the intervention is introduced. The 2x2 design entails the assignment of participants to an intervention group and a waitlist control group. Participants will be matched across the two groups using the scores and information obtained during the screening process. Participants assigned to the EMIT group will receive the intervention immediately. Participants in the waitlist control group will begin participating in the research sessions when their assigned matched study participant has completed his/her participation in the sessions.

Sessions will be conducted in a room with minimal distractions containing a table, two chairs, a timer, and 3-4 highly preferred (HP) items identified by preference assessments.A minimum of 10 sessions per week will be distributed across at least 3 days per week. Each session will take 20-25 min to conduct. Each session will consist of the instructor presenting imitation training trials and probe trials. During baseline sessions, participants will not receive rewards for correct imitation. During training, participants will receive rewards for correct imitation and evidence-based teaching strategies that are commonly employed in ABA therapy will be used to help participants learn how to imitate new responses. Novel probe trials will be embedded throughout all sessions to continuously evaluate the emergence of generalized imitation. When participants can consistently imitate novel probe responses, the intervention will conclude.

Inclusion Criteria: Diagnosis of autism spectrum disorder Age 5 or under Maximum score of 40 on Motor Vocal Imitation Assessment Maximum raw score of 36 on Peabody Picture Vocabulary Test-5 Demonstrates identity matching above chance levels Exclusion Criteria: Diagnosis of intellectual disability Diagnosis of certain conditions that impact motor function (e.g., cerebral palsy) Visual impairment Hearing impairment