Part 1: How PECS helped condition a better token.

Token economies are one of the most common techniques used in behavioral interventions. A literature search using the term, “token economy” will result in over a thousand hits. Talk to practitioners about tokens and most, if not all, will speak with familiarity. In behavioral vernacular, the term “token” is used that implies a specific procedure – e.g., “Let’s starts tokens” implying, let’s begin a token conditioning procedure. Often, there might not be further discussion – that implementing a token economy is so common that all practitioners are qualified to begin these protocols. Alternatively, the assumption might be that a token economy is such a simplistic intervention that simple variations in the protocol are not obstacles to effectiveness.

A conversation with the school’s Clinical Supervisors begged the simple question, how should we start. We discussed several procedural alternatives for how a neutral item, such as a poker chip, or penny, or sticker might acquire reinforcing properties and quickly realized we were not sure why one procedure would be better than any other. So, as good behavior analysts, we looked to the published literature. And, as stated before, found a list of articles numbering over a thousand. Surprisingly, we could not, however, find articles that would help us with our question: What are the factors that influence conditioning of a secondary/generalized reinforcer (with people with autism)? Despite the numerous articles, studies typically focused on some other behavior change as a function of token reinforcement. We were hard-pressed to find those articles that actually described the conditioning process of the token itself. So, we looked elsewhere.

We asked, what kind of behavior do we want to see when using an effective secondary/generalized reinforcer? We thought of the typical example - $$.

1. We exchange money for other things: If we had an endless supply of money, we would exchange it a lot!

2. We like money: Under normal circumstances, most of us would choose money over other tangible items if given a choice.

3. We don’t mistreat our money: It’s valuable, we collect it, and save it for future use.

We felt that these observations matched results another procedure where a conditioning process occurred quickly and effectively – the Picture Exchange Communication System (aka PECS). For the non-communicating person, the initial picture exchange training protocol quickly leads to exchange behavior – giving a picture to gain preferred or reinforcing items. Typically, the initial training protocol is so effective that independent exchanges can occur within one or a few training sessions. Quickly, exchange behavior occurs at such a high-frequency that some practitioners even consider this high-rate, early communication behavior as a problematic!

The picture exchange training protocol was the perfect model for a token-conditioning protocol. Would it work? The short answer – Yes, as demonstrated by Weppler, Pichardo, Fennell, & Lee in a poster presented at the 2009 ABAI convention in Phoenix. Procedures and results in next month's blog.