My entire career, I've been a relationship kind of a guy. More specifically, I've favored taking a relationship-based approach to helping children with autism. The belief was (and is), that when an autistic person pulls away into his own world, this is something to be worked with rather than battled. Allowing children to pull away gives them an opportunity to establish some level of control, take a break, and /or follow a plan. All of these have value if you believe, as I do, the relationship an autistic person has with the others in his life will be a primary motivator for learning skills and, more importantly, consistent and deep interaction.

I still strongly believe in relationship-based approach. That said, in the past six months, I've come to see strong value in table work and trials (traits from the pretty-much-opposite method of Applied Behavioral Analysis).

Even now, looking back, it makes sense: taking the best practices from multiple approaches works in most fields, why wouldn't it work in autism? It sounds great on paper, but there can be huge resistance from professionals who have been practicing one particular method for their entire careers (as I have).