Intervention in the Community for Toddlers with ASD: Paraprofessionals’ Ratings of Implementation Complexity and the Association with Observed Intervention Implementation

Background: Treatment fidelity is a concern across intervention studies for children with autism both in clinic and when transferring practices into the community (Dingfelder and Mandell, 2011). To enhance researchers’ ability to support community implementation, it is necessary to better understand factors that may influence intervention fidelity. The present examination looks at paraprofessionals teaching assistants (TAs) observed implementation scores and their self-report of perceived complexity to deliver the intervention including effort, experience, as well as comfort and confidence in implementing the intervention.

Objectives: To determine whether there is an association between TAs’ intervention implementation fidelity and their self-reports of the perceived complexity of the intervention.

Methods: The analysis includes 43 children who were randomized to receive the JASPER intervention (Kasari et al, 2006; 2008) as part of a larger intervention trial. Eighteen TAs delivered the JASPER intervention. On average TAs were 30.25 years old and worked at the center for 2.27 years. All but one of the TAs were ethnic minorities. Ten-minute TA-child play interactions were videotaped at treatment exit and coded for strategy use (implementation). Implementation codes include subscale scores (i.e., basic strategies, setting up the environment, following the child’s lead, play routines, joint attention/requesting skills, and language). TA Diaries (adapted from Kasari et al., 2003) were completed at study entry. TAs rated 6 questions using a 1-5 scale (1=strongly disagree, 5=strongly agree) to represent their experience in implementation (e.g., strategies feel new/complicated, goes beyond their comfort zone, requires extra work/time). These items were summed to report total perceived complexity, the degree to which TAs felt the intervention was challenging to implement.

Results: Average total diary scores were 17.07 (SD=4.31) out of a total possible score of 30. A Pearson bivariate correlation coefficient was computed to determine the association between implementation and complexity rating scores. There was a positive correlation between the fidelity subscale of following the child’s lead and total perceived complexity rating (r= 0.333, n= 43, p=0.029). There were also a positive correlation between the fidelity subscale of setting up the environment and total difficulty score (r=0.354, n=43, p=0.020). TAs’ average implementation of setting up the environment in the sample was 90.02% (SD=10.78%), and average implementation of following the child’s lead was 85.72% (SD=11.31%).

Conclusions: TAs’ initial perception of implementation complexity may be linked to higher quality intervention implementation in the particular areas of setting up the environment and following the child’s lead. In this intervention, these strategies are foundational, providing a framework to layer on more strategies and learning opportunities. Awareness of the time and work required to implement unfamiliar environment and following strategies is correlated with implementing these strategies more effectively. It is possible that focusing more on the complexities and newness of an intervention translates to better foundational practices because these strategies are the first thing that the TAs do in implementation before incorporating higher level strategies.