Variability as an Operant Dimension
Behavior analysis has traditionally focused on establishing reliable, consistent responding. However, research over the past several decades has demonstrated that variability in behavior can itself be an operant that is sensitive to reinforcement contingencies. Just as rate, duration, and intensity of behavior can be shaped through reinforcement, so can the degree of variability in a behavioral repertoire. This finding has important theoretical and practical implications.
Allen Neuringer's research, beginning in the 1980s and continuing for decades, provided the foundational evidence that variability is directly reinforceable. Neuringer demonstrated that organisms could learn to respond variably when variability was the explicit criterion for reinforcement, and that this variability was not simply a byproduct of extinction or other processes but was directly controlled by its consequences.
Lag Schedules of Reinforcement
What Is a Lag Schedule?
A lag schedule of reinforcement specifies that a response is reinforced only if it differs from the previous n responses. The number following "lag" indicates how many preceding responses the current response must differ from. For example:
- Lag 1: The response is reinforced if it differs from the immediately preceding response
- Lag 2: The response is reinforced if it differs from the two most recent responses
- Lag 3: The response is reinforced if it differs from the three most recent responses
As the lag value increases, greater variability is required to meet the reinforcement criterion. A Lag 1 schedule requires only that the current response differ from the last one, which can be achieved by simply alternating between two responses. A Lag 5 schedule requires that the current response differ from all of the last five responses, demanding a much more varied repertoire.
Research Findings
Neuringer and colleagues demonstrated several key findings about lag schedules:
- Variability increases systematically as the lag requirement increases
- The variability produced by lag schedules is genuine operant variability, not random or extinction-induced variability
- Lag schedule training can produce levels of variability that approximate mathematical randomness
- Variability trained under lag schedules can generalize to novel response topographies and new contexts
- Both human and nonhuman organisms respond to lag contingencies
Clinical Applications
Promoting Creative Responding
One of the most exciting clinical applications of lag schedules is promoting creative or flexible responding in individuals who tend toward rigid, repetitive behavior patterns. Many individuals with autism spectrum disorder demonstrate restricted and repetitive behavioral repertoires. Lag schedules have been used to increase the variety of play behaviors, verbal responses, social initiations, and other skills where repetitive responding limits functional outcomes.
For example, a child who always builds the same structure with blocks can be placed on a lag schedule where reinforcement is provided only for block constructions that differ from the previous one (or the previous several). Over time, the child's building repertoire becomes more varied and creative. Lee, McComas, and Jawor (2002) demonstrated that lag schedules of reinforcement effectively increased variability in children's toy play.
Increasing Verbal Variability
Lag schedules have been applied to increase the variety of verbal responses. A child who always uses the same sentence structure to request items can be reinforced for using different sentence frames (e.g., "I want the ___," "Can I have the ___," "May I please get the ___"). By reinforcing only responses that differ from recent ones, the clinician builds a more flexible and socially typical verbal repertoire.
Susa and Schlinger (2012) and others have demonstrated that lag schedules can increase the variability of intraverbal responses, mands, and other verbal operants. This application is particularly relevant for individuals whose verbal behavior is overly scripted or echolalic.
Social Skills Training
Variability training through lag schedules has also been applied to social skills. An individual who always initiates social interactions in the same way (e.g., always saying "Hi, what's your name?") can be taught to use varied greetings and conversation openers through lag contingencies. This produces more natural-sounding social behavior and reduces the impression of scripted responding.
Distinguishing Operant Variability from Other Sources
Not all behavioral variability is operant variability. It is important to distinguish variability that is directly reinforced from variability that results from other processes:
- Extinction-induced variability: When a previously reinforced response is placed on extinction, the organism may try novel responses. This variability is a byproduct of the extinction process, not a directly reinforced operant.
- Adventitious variability: Random fluctuations in behavior that are not systematically related to any contingency
- Resurgence: The reappearance of a previously reinforced behavior during extinction of a more recently reinforced behavior
- Operant variability: Variability that is directly maintained by a reinforcement contingency (such as a lag schedule) that requires variation for reinforcement to occur
Theoretical Implications
The finding that variability is an operant dimension has significant theoretical implications for behavior analysis. It demonstrates that even the degree of predictability in behavior is controlled by environmental contingencies. This challenges the misconception that behavior analysis can only account for repetitive, machine-like behavior. In fact, creative, novel, and flexible responding can all be understood within an operant framework when the reinforcement contingencies specify variability as the criterion.
Neuringer's work also has implications for understanding creativity more broadly. From a behavior-analytic perspective, creative behavior can be understood as behavior that is both varied (novel or unusual) and effective (reinforced by the environment). Lag schedules directly train the variability component, while other contingencies can shape the effectiveness component.
BCBA Exam Preparation
Behavioral variability questions on the BCBA exam may ask you to:
- Define a lag schedule and explain how it promotes variability
- Distinguish operant variability from extinction-induced variability
- Identify clinical scenarios where lag schedules would be appropriate
- Explain how increasing lag values produce increasing variability
- Recognize that variability can be directly reinforced as an operant dimension
Key points to remember: Lag 1 is the simplest level, requiring only that the current response differ from the last one. Higher lag values demand more varied responding. The variability produced is genuine operant behavior, not random variation. Neuringer is the foundational researcher in this area. Lag schedules have practical applications for individuals with restricted behavioral repertoires, and the concept bridges behavior analysis with the study of creativity and flexible problem solving.