Imitation in Improvisational Music Therapy Supports Engagement in Children Autism Spectrum Disorder

Background:

Imitation plays a crucial role in learning to communicate, and allows individuals to engage in reciprocal communication. Previous research by has suggested that exact imitation is more effective than contingent responding at improving social interaction in children with autism. In addition, many naturalistic treatments for autism include an imitation phase at the beginning of the treatment to foster engagement and joint attention. For some treatments, the imitation is exact while others have suggested that elaborative imitation can serve both to build rapport with the child as well as provide a means of social-interaction. This study is the first study to test the specific technique of imitation in improvised music therapy (IMT) and is the first to test the relative effectiveness of exact imitation versus elaborative imitation, and contingent responding.

Objectives:

This study tests the relative effectiveness of three treatment strategies used during IMT for children on the autism spectrum.

 Methods:

7 preschoolers with autism participated in the study. A single-subject multiple treatment design tested the relative effectiveness of three IMT treatment techniques: 1) exact imitation (ExI) in which the therapist imitated the rhythm or melody of each child’s turn, 2) imitation with elaboration (IE) in which the therapist first imitated the child directly and then added to or elaborated on the child’s turn, and 3) contingent response/accompaniment (CRA) in which the therapist played music to match the tempo, intensity, or mood of the child’s turn. Children participated in three 30-minute therapy sessions of each condition. The order of conditions was randomized for each participant. Sessions were videotaped and later coded for treatment fidelity of each of the therapist’s turns. Relative effectiveness of the treatments was assessed by measuring child engagement behaviors using the 5 non-symbol infused engagement states from Adamson et al.’s (2000) engagement scale.

 Results:

The results suggest that ExI (.22) and IE (.23) resulted in the highest median rates of positive engagement behaviors, and the lowest median rates (.06 and .09 respectively) of unengaged behaviors. CRA resulted in the lowest median rate of engaged behaviors (.14) and highest median rate of unengaged behaviors (.11). Scrutiny of individual participant’s scores, however, suggests that certain treatment conditions are better at encouraging certain types of engagement. EI almost uniformly (6/7 participants) resulted in the worst coordinated joint attention while CRA resulted in the best coordinated joint attention in 5 out of 7 participants. Supported joint attention, on the other hand, was best in the EI.

 Conclusions:

This the first study to compare the effectiveness of individual treatment strategies used in IMT. The results suggest that techniques that involve imitation (either exact or elaborated) may be best at encouraging positive engagement states in children on the autism spectrum during improvised music therapy. We note that as a single-subject study, this is only a first foray into the topic and further work needs to be done with a more rigorous experimental design.