Learners with autism spectrum disorders (ASD) have diverse and unique needs. They struggle with attention, communication, cognitive delays, and behavioral control. Their learning and behavioral challenges are highly idiosyncratic, and educating students with ASD requires a fully individualized, intensive, and comprehensive educational program.
The most effective intervention for educating students with ASD is Applied Behavior Analysis (ABA), which has been documented to be extremely effective in building a wide variety of skills. ABA intervention is evidence-based and uses empirically validated instructional strategies, develops precise educational goals, and uses ongoing collection of data to monitor student progress and make data-based decisions. ABA intervention has been criticized as educating students with ASD in a robotic fashion, but it is actually a dynamic and flexible educational programming approach, enabling the practitioner to not only develop carefully constructed incremental programs for skill acquisition, but also to assess the functions of challenging behaviors and develop behavior intervention plans matched to function, analyze and program the generalization of skills to the natural environment.
A combination of discrete trial instruction and naturalistic instructional approaches is used to build skills across all curricular areas. There is generally a strong focus on providing reinforcement to build behaviors, on prompting to facilitate skill acquisition, and on assessing the transfer of skills. Quality ABA programs focus on providing a high number of learning opportunities and on capitalizing on naturally occurring opportunities for instruction.
Many programs that utilize ABA also incorporate interventions from other disciplines, such as speech and occupational therapy to address communication and sensory issues. ABA can be used in combination with these approaches to assist practitioners first in gaining behavioral control so they can institute their interventions more effectively. ABA can also be helpful in refining interventions by developing operationally defined targets, in standardizing some interventions within these approaches to assist in tracking progress, and in evaluating the impact of these methods for particular learners.
Outcome of ABA When Educating Students With ASD
As mentioned above, there is a substantial data base on outcome with behavior analytic intervention. Lovaas (1987) compared a group of children under age 4 who received 40 hours of intervention per week for 2 or more years with groups of children who received fewer hours of intervention. Almost half of the children in the intensive intervention group were able to be placed unassisted in regular education classes and achieved IQ’s in the average range. Other researchers have replicated that early intensive behavioral intervention results in significant gains for some children (e.g., Green, Brennan, & Fein, 2002; Smith, 1999). From a clinical perspective, it is important to note that more research is still needed to completely understand the effective elements of instruction. Furthermore, outcome remains highly variable, and reliable predictors of outcome have not been clearly identified.
One of the central issues for both parents and professionals is the identification of quality indicators of programs serving individuals with autism. Parents seek information on this as they seek to locate appropriate programs that will maximize their child’s skills. Professionals seek to identify these variables to develop, refine, and improve their educational services.
Most programs utilizing ABA focus on using a combination of formal and naturalistic strategies. Discrete trial instruction has been effective in educating students with ASD a wide variety of core skills in a structured, formalized context. Elements of its effective use include errorless learning procedures (e.g., Etzel & LeBlanc, 1979; Lancioni & Smeets, 1986; Terrace, 1963; Touchette & Howard, 1984) and task variation and interspersal (e.g., Dunlap, 1984; Mace, Hock, Lalli, West, Belfiore, Pinter, & Brown, 1988; Winterling, Dunlap, & O’Neill, 1987; Zarcone, Iwata, Hughes, & Vollmer, 1993). In addition, discrete trial teaching is much more effective if utilized with strategies for effective generalization to the natural environment (Smith, McAdam, & Napolitano, 2007; Stokes & Baer, 1977). Strategies that facilitate transfer to the natural environment include the use of varied phrases in instruction and the use of a naturalistic tone of voice in delivering instructions.
Other elements of effective instruction within formal discrete trials include attention to pacing of instruction and to the provision of frequent and motivating rewards. In pacing, the goal is to maximize the number of opportunities there are for the student to learn in an instructional session. Fast-paced instruction increases learning opportunities, increases access to reinforcers, ensures high rates of engagement, and prevents inappropriate and off-task behaviors. Even when formal instruction is delivered in groups, it is important to attend to the pace of instruction, as faster pace facilitates attention, engagement, and appropriate behavior.
Programs utilizing ABA are committed to taking data, not just for accountability, but for quality control and for decision-making purposes. From a quality control perspective, the collection of data provides information on the integrity of treatment (i.e., is treatment being implemented as designed?) For decision-making, data is collected and reviewed and analyzed on a daily basis to provide information on rate of acquisition, error patterns, the need for additional prompts, and recommendations to alterations in treatment.
Data were collected across a wide variety of learners and teachers in our school (15 learners, 12 staff members). The data allow for the evaluation of treatment integrity along some critical dimensions (e.g., are there at least 5 instructions delivered per minute in intensive individualized instruction? ; are at least 4 reinforcers delivered per minute in intensive individualized instruction?; are there at least 4 instructions delivered per minute in group instruction?) Results of the analyses indicated that instructors delivered 6.5 instructions per minute in individual instruction, and 4.8 times per minute in group instruction. Reinforcers were delivered 4.5 times per minute in individual instruction, and 3.1 times per minute in group instruction. All of these numbers fall within the indicators of high quality instruction, and reflect intensive programming and high rates of student engagement.
Other indicators of quality programs involve aspects of program organization, environmental set-up, or classroom management. These variables include ensuring staff readiness for instruction and involving the learner in decisions about the sequence or types of activities selected. Data were collected on several aspects of these antecedent/preventive strategies to assess whether the environment was maximally conducive to learning. Data were collected in 16 observations, across 8 learners and 10 staff members, and indicated that staff members were maximally prepared for instruction 97% of the time, and that choices were provided to students in 96% of potential opportunities. These data indicate good use of strategies associated with lower levels of challenging behaviors.
In addition to maximizing formal instruction and ensuring an optimal environment, there is also a need for naturalistic instruction. While the formality of discrete trial instruction builds responsivity, naturalistic strategies build initiation. Initiation skills are critically important for navigating environments independently, and for reducing the need for instructors to anticipate the learner’s needs. One naturalistic approach is incidental teaching, in which the learner initiates a request or a conversation about a particular item or topic. The teacher prompts an elaboration of that initiation, and the learner’s more elaborate communication results in immediate access to the desired item (Fenske, Krantz, & McClannahan, 2001). One of the most important aspects is that the learner is leading the teaching interaction, as his or her interests create the opportunity for the instruction (Fenske, Krantz, & McClannahan, 2001). Incidental teaching is an excellent way to increase initiation and spontaneity.
Other naturalistic methodologies within ABA have also emphasized the themes of initiation and learner interests. Pivotal Response Training (PRT) and Natural Language Paradigm (NLP) emphasize using high interest and motivating materials, teaching in natural situations, and capitalizing on the child’s interests to target deficits in language (Koegel & Koegel, 2005; Koegel, Koegel, & Surrat, 1992; Koegel, O’Dell, & Koegel, 1987; Laski, Charlop, & Schreibman, 1988). Natural Environment Training (NET; Sundberg & Partington, 1998), like NLP and PRT, focuses on the use of intrinsically motivating materials and on following the child’s lead in language instruction. NET also uses Skinner’s Verbal Behavior language classification system/analytical tool to guide language instruction (Skinner, 1957).
Elements of naturalistic instruction include providing an environment in which learners request spontaneously. In addition, naturalistic teaching interactions that facilitate rapport building, and pair the instructor with reinforcement, are also desirable. The following data were collected in 12 samples of 8 children with 10 instructors. Mands occurred an average of 1.4 times per minute, indicating a high level of initiation training. Instructors also delivered an average of 6 non-contingent reinforcing statements in 5 minutes, demonstrating good adherence to pairing (rapport-building) procedures.
Quality programs for children with autism spectrum disorders are essential to success. Children with autism who receive non-specialized and ineffective instruction do not make significant gains, and have significantly less positive educational outcomes and achievements. It is imperative that parents understand the hallmark characteristics of quality programs, as educational placement directly affects ultimate outcomes. Programs delivering high quality services operationally define their instructional goals, use empirically verified instructional strategies, and rely on ABA teaching technology to provide both formal and naturalistic instruction. High quality programs also ensure the fidelity of the implementation of these approaches, and collect data to guide the treatment decisions made for each learner.
Mary Jane Weiss, PhD, BCBA, is Clinical Consultant, Cecelia M. McCarton, MD, is the Founder and CEO, Ivy Feldman, PhD, BCBA, is Educational Director, and Jackie Hickey, M.S, BCBA is Associate Educational Director at The McCarton Foundation.