Perkins School for the Blind Transition Center

Improving Communication Skills: Using Behavior Analytic Science Effectively

Learners with Autism Spectrum Disorder (ASD) often face significant challenges learning socially appropriate and effective communication skills. Because these challenges are present across all ages and stages of development, programming targets can range from communicating basic wants and needs to navigating complex conversations. Practitioners must be skilled in creating a variety of interventions that best address these diverse communication needs.

Cresse Morrell, MS, BCBA, LBA (CT)

Cresse Morrell, MS, BCBA, LBA (CT)
Vice President of Clinical Operations

Another factor affecting service delivery for the ASD community is the workforce of practitioners. Recently certified practicing behavior analysts may not have had a wide range of programming experience. Some of this growth can be partially attributed to the overwhelming need for service in the ASD community. As such, it is increasingly important to ensure that interventions are developed and implemented with fidelity and integrity. A parsimonious approach to achieving those standards of practice is to incorporate the seven dimensions of Behavior Analysis as described by Baer et al. (1968). Since some of those dimensions may prove to be more challenging or complex to weave into practice, the use of a set of guided questions specific to each dimension can assist in isolating the most relevant components.

Developing appropriate and effective programs requires the practitioner to carefully construct the foundation and framework upon which the program will start and progress to mastery. As referenced above, a good starting point may include having the team answer a set of guiding questions. What follows are some important components to consider during the process to developing, implementing, and monitoring effective programs.

  • Where will the program be taught? Why?
  • Where will the skill be used once the program meets criteria?
  • How will you know that the skill is being used in the appropriate place(s)?
  • How do typical peers use the skill? How will you ensure that your student does the same?
  • How will you take data? Frequency, independence, percent correct?
  • How will you define final criteria for use of skill?
  • How will you ensure that the skill is under the appropriate stimulus control?
  • How will you define the behavior?
  • How will you describe the procedure? Make sure you include all potential situations that could occur.
  • How will you decide if the behaviors taught are socially significant for that student?
  • How will you ensure that the appropriate function is identified?
  • Include use of reinforcement, punishment, use of motivating operations, shaping, fading, thinning of schedules of reinforcement, chaining.
  • How often will you measure the behavior?
  • How will you determine criteria for mastery?
  • How will you know when to make a change?
  • How will you ensure that everyone is measuring the same behavior?

Many of the above questions are standard to behavior analytic programming; however, several warrant a more in-depth discussion. What follows are specific considerations across some of the questions using the example of teaching manding (requesting) to a learner who is beginning to communicate using vocal speech.

Effective communicators have the ability to ask for wants and needs in any environment (home, school, the mall, an airport, or a brand new environment). In order to ensure generalization occurs, programs should be taught in every environment the learner visits so that responses are robust and can be used across all people and places. Communication skills cannot be limited under the stimulus control of a single environment.

Another measure of successful intervention is the production of socially important effects. This requires the practitioner to understand both topography (what the behavior looks like) and function (the purpose the behavior serves). While topography can and should be shaped to what is socially acceptable, addressing the function of communication is the first and most important factor. Delivering a communication program in a discrete trial format may not allow for generality or consider function. For example, vocally or non-vocally responding with the word “cookie” in the presence of a cookie is only functioning as a mand if there is an existing motivating operation in place. Merely repeating an instructor’s vocal prompt does not mean requesting is being taught. Practitioners must ensure that responding is under the correct stimulus control and that programming is appropriately matched to that function. Careful consideration must be given to the choice of items specific to the amount of the item or amount of time spent engaging with the item, how much and how recently the learner has accessed the item, and how quickly the learner receives the item or access to the item after manding. Simply put, if a person doesn’t want an item, you do not have an opportunity for teaching manding.

An important element to consider when designing contingencies for manding programs is how teaching strategies will be implemented. Reinforcement contingencies for manding for an item should produce delivery of that item. Praise and tokens are not functional reinforcers for manding for a cookie. The reinforcer for manding for a cookie is receiving a cookie. It is also of great importance to fade prompts immediately. When using vocal prompting protocols, fading procedures must be in place. If a vocal prompt is controlling the learner’s vocal response, that relation is not functional. Moreover, if those prompts are not faded carefully and quickly, development of a completely separate relation can occur. It is probable that either tacting or simply a stimulus (the vocal prompt) and a response (the word cookie) are the results of ineffective prompt fading in this example.

When analyzing data for manding programs, practitioners should evaluate data from teaching interactions across the learner’s day and not just in sessions. Not only does this provide a more robust data set, the process also adds additional assurance that the program is being taught across environments. Analysis of data includes looking at criteria for mastery. Peer behavior should be one of the variables considered in determining mastery criteria. If the goal of the manding program is to teach the learner to use the skill optimally, the frequency with which it occurs should match peer behavior as closely as possible.

While the above illustration is only a snapshot of the depth and careful consideration that diligent programming requires, the use of the seven dimensions of Behavior Analysis when constructing intervention can address some common pitfalls. Ensuring that each of the seven dimensions are included in communication programming creates interventions that have the best possible chance for success and meet current recommended practice. Working to change human behavior requires critical and analytic skills as well as an ability to accurately and effectively provide rationale for interventions. Early career practitioners may find that the use of a set of guided questions can help focus intervention for ideal programming results.

I can be reached at and Milestones website can be found at


Baer, D.M., Wolf, M.M., & Risley, T.R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1, 91-97.

Have a Comment?