Applied Behavior Analysis: Contingencies, Function, and Single-Case Design

Abstract

Applied behavior analysis (ABA) is a form of behavior therapy that applies the principles of operant conditioning to socially important behavior, changing environmental antecedents and consequences to produce durable, measurable change. Its 1968 founding paper defined seven dimensions separating it from mere behavior modification. Two contributions distinguish it methodologically: the experimental functional analysis, which isolates the environmental function a problem behavior serves before treatment is chosen, and the single-case experimental design, which demonstrates a functional relation within one participant by reversing or staggering the intervention. This article states the science ABA applies, the three-term contingency at its core, the logic of functional analysis and single-case measurement, and the contested evidence for its most publicized application, early intervention in autism. Three interactive demonstrations let the reader manipulate a reinforcement contingency, run a functional analysis, and construct a reversal design.

Keywords: applied behavior analysis, operant conditioning, reinforcement, functional analysis, single-case design

Applied behavior analysis is the applied wing of the experimental analysis of behavior, the science of operant conditioning that B. F. Skinner built from the 1930s onward (Skinner, 1953). Where the basic science establishes lawful relations between behavior and its consequences under controlled laboratory conditions, ABA takes those relations into homes, classrooms, clinics, and workplaces and asks whether they can be used to change behavior that matters to the people living it. The discipline dates its own founding to 1968, when Donald Baer, Montrose Wolf, and Todd Risley published a set of criteria — applied, behavioral, analytic, technological, conceptually systematic, effective, and generality-producing — that separated a genuine analysis of socially important behavior from the looser practice of behavior modification (Baer et al., 1968). Those seven dimensions remain the field's self-definition, revisited but not replaced two decades later (Baer et al., 1987).

Key Takeaways

ABA applies operant principles — reinforcement, punishment, extinction, stimulus control — to behavior of social importance, held to seven defining dimensions since 1968.
Its unit of analysis is the three-term contingency: antecedent, behavior, consequence. Behavior is changed by changing what precedes and what follows it, not by inferring an inner cause.
The experimental functional analysis identifies the environmental function a problem behavior serves — attention, escape, tangible access, or automatic reinforcement — before a treatment is selected.
ABA demonstrates effects through single-case experimental designs, in which each participant serves as their own control across repeatedly reversed or staggered phases.
Its evidence base for early autism intervention is substantial but contested, and a neurodiversity-informed critique of intensity and compliance goals has reshaped contemporary practice.

What Applied Behavior Analysis Is

Applied behavior analysis is defined by its subject matter and its method together, not by either alone. Its subject matter is socially significant behavior — skills a person needs and problems that impair a life — and its method is the direct manipulation of the environmental variables of which that behavior is a function. The field takes behavior itself, not a hypothesized inner state, as the thing to be explained and changed; a tantrum is analyzed as a response with an environmental history and an environmental function, not as a symptom of an underlying disturbance to be interpreted. This is the applied expression of Skinner's radical behaviorism, which treated private events as behavior subject to the same principles rather than as causes standing outside the behavioral stream (Skinner, 1953). Skinner extended the same functional account to language itself, analyzing verbal responses by the contingencies that occasion and maintain them rather than by reference to meaning or intention (Skinner, 1957); that framework underwrites ABA's approach to teaching communication as a repertoire of functional verbal operants.

MeSH classifies applied behavior analysis under behavior therapy (descriptor D000072357, tree F04.754.137.131), and the placement is accurate: ABA is one of the behavior therapies, sharing their reliance on learning principles rather than insight or interpretation. What sets it apart within that family is methodological rigor imported wholesale from the operant laboratory — continuous direct measurement of behavior, the three-term contingency as the unit of analysis, and single-case experimental control — rather than any distinctive clinical technique. A token economy, a program of differential reinforcement, or a graduated extinction procedure is behavior therapy done to the analytic standard ABA sets for itself.

The Seven Dimensions

The 1968 founding paper is organized around seven words, each naming a property a project must have to count as applied behavior analysis (Baer et al., 1968). Applied means the behavior studied is socially important to the participant, not merely convenient to study. Behavioral means what is measured is behavior itself, reliably and directly, rather than a verbal report about it. Analytic means the analyst can demonstrate control — can show that manipulating the intervention reliably turns the behavior on and off. Technological means the procedures are described precisely enough that another practitioner could replicate them. Conceptually systematic means the procedures are derived from and described in terms of basic behavioral principles, so the field accumulates a coherent technology rather than a bag of tricks. Effective means the change is large enough to matter practically, not merely to reach statistical significance. And generality means the change endures across time, settings, and behaviors once the program ends.

The most consequential of the seven for daily practice is the analytic dimension, because it commits ABA to demonstrating, not assuming, that its intervention is responsible for the change observed. The revisiting paper two decades later added little to the list but sharpened one point: generality is not a happy accident to be hoped for but an outcome that must itself be programmed, through techniques such as training loosely, using natural maintaining contingencies, and reinforcing across multiple exemplars (Baer et al., 1987). A later, equally durable addition came from Wolf, who argued that effectiveness must be judged partly by the people it affects — that a program's goals, procedures, and results should be socially valid in the eyes of consumers, not only the analyst (Wolf, 1978).

Figure 1

The Three-Term Contingency

Note. Behavior is selected by its consequences: a consequence that raises the future probability of the response is a reinforcer, one that lowers it a punisher, and the antecedent sets the occasion on which the contingency operates. Original schematic.

Reinforcement and the Three-Term Contingency

The unit of analysis in ABA is the three-term contingency: an antecedent that sets the occasion, the behavior itself, and the consequence that follows. Operant principles are statements about the third term. A reinforcer is any consequence that increases the future probability of the behavior it follows; a punisher is any consequence that decreases it. The definitions are functional, not physical — a consequence is a reinforcer only if it in fact raises the response rate, which is why a scolding that a child finds attention-worthy can reinforce the very behavior it was meant to suppress. Crossing the direction of the behavioral effect with whether a stimulus is added or removed yields the familiar four-cell operant matrix: positive and negative reinforcement both strengthen behavior, by adding a stimulus or removing one; positive and negative punishment both weaken it, again by adding or removing.

The word negative in this scheme means subtracted, never bad — negative reinforcement strengthens behavior by removing an aversive stimulus, as when fastening a seatbelt terminates a warning chime. Confusing negative reinforcement with punishment is the single most common error in the whole vocabulary, and it is a lexical problem rather than a conceptual one. The demonstration below lets the reader select each cell of the matrix and watch its characteristic effect on a response rate over successive sessions; the point it makes is that the label attaches to the effect on behavior, read off the data, not to the practitioner's intention.

The rate and pattern of responding depend not only on which contingency operates but on the schedule by which reinforcement is delivered — continuously, after a fixed or variable number of responses, or after a fixed or variable interval of time. The exhaustive mapping of how each schedule shapes a characteristic, orderly pattern of responding, worked out in the operant laboratory, is the empirical foundation ABA inherits for arranging reinforcement in applied settings (Ferster & Skinner, 1957).

The operant matrix and its effect on behavior

Choose a contingency and watch its characteristic effect on the response rate over twelve sessions. The label names what happens to the behavior, not the practitioner’s intention: both kinds of reinforcement strengthen it, both kinds of punishment weaken it.

Positive reinforcement (add an appetitive stimulus) raises the response rate from a baseline of 5/min to 13.8/min by session 12. Whether the stimulus is added or removed, it is the direction of the change in behavior that defines the contingency as reinforcement or punishment.

Figure. Deterministic exponential approach to a contingency-set asymptote; no data are sampled. Original illustration.

The Functional Analysis of Behavior

ABA's most distinctive methodological contribution is the experimental functional analysis, introduced in the form still used today by Brian Iwata and colleagues, first as a 1982 monograph and formalized in its 1994 reprinting (Iwata et al., 1994). The premise is that the same problem behavior — self-injury, aggression, disruption — can be maintained by entirely different environmental functions in different people, and that a treatment which ignores the function is as likely to worsen the behavior as to help. A behavior maintained by adult attention will be strengthened, not extinguished, by an adult who reliably attends to it in order to intervene.

The analysis tests the candidate functions directly by arranging brief, alternating conditions, each of which delivers one class of consequence contingent on the target behavior: a social attention condition, a demand condition that permits escape from a task, a tangible condition that delivers a preferred item, an alone condition that removes all social consequences, and a play control condition rich in noncontingent reinforcement. The behavior rises in the condition whose consequence maintains it and stays low in the others; the differentiated pattern names the function. Once the function is known, treatment follows from it — teaching a functionally equivalent replacement response, such as a communicative request that earns the same consequence more efficiently, and withholding that consequence for the problem behavior. The demonstration below runs an idealized functional analysis: the reader selects which function is operating and observes the differentiated response rates the analysis would reveal.

Reading a functional analysis

Select the function that truly maintains the behavior, then read the differentiated pattern the analysis would produce. The condition standing well above the play control names the function; the others hover near control.

With the true function set to Social attention, the behavior runs at 8.4/min in the Attention condition — 7.0× the play-control rate of 1.2/min — while the remaining conditions stay near control. The differentiated pattern identifies the maintaining consequence, and treatment then teaches a functionally equivalent replacement that earns the same consequence.

Figure. Idealized analog functional-analysis outcomes; rates are fixed per function, not sampled. Original illustration after the design of Iwata et al.

Measurement and Single-Case Design

Because ABA commits to demonstrating experimental control, and because it works with individuals rather than groups, it developed its own experimental methodology: the single-case (or single-subject) design, in which each participant is their own control. The logic is repeated demonstration of an effect within one person over time. In the reversal or ABAB design, behavior is measured under a baseline condition (A), then under the intervention (B), then the intervention is withdrawn to return to baseline (A), then reinstated (B). If the behavior tracks the condition — changing when treatment is introduced, reverting when it is removed, changing again when it returns — a functional relation is established, because chance and maturation cannot readily produce a pattern that turns on and off in step with the manipulation.

Reversal is not always ethical or possible: one does not withdraw a treatment that has stopped self-injury, and some learned skills do not reverse. For those cases the field uses the multiple-baseline design, which staggers the introduction of the intervention across several behaviors, settings, or participants and shows that each changes only when, and not before, the intervention reaches it. Measurement throughout is direct and continuous — frequency, duration, latency, or interval recording of the behavior itself — and results are displayed on a line graph and judged by visual analysis of level, trend, and variability across phases rather than by a significance test. The demonstration below lets the reader build an ABAB design phase by phase and see how the reversal exposes, or fails to expose, experimental control.

Building an ABAB reversal design

Reveal the design one phase at a time. If the behavior drops when the intervention is introduced, returns toward baseline when it is withdrawn, and drops again when it is reinstated, chance and maturation cannot explain the pattern — a functional relation is demonstrated.

Showing 1 of 4 phases

3 phases still hidden. With only 1 phase revealed, the effect is not yet demonstrated — a single baseline shows only how the behavior looks untreated. Add the remaining phases to complete the reversal.

Figure. Fixed per-session values for each phase; nothing is sampled. Original illustration.

Function	What maintains the behavior	Function-based treatment
Social attention	Positive reinforcement by others attending to it	Teach a request for attention; withhold attention for the problem behavior
Escape / demand	Negative reinforcement by termination of a task	Teach a break request; do not allow escape to follow the behavior
Tangible access	Positive reinforcement by a preferred item or activity	Teach an item request; deliver the item only for the request
Automatic	Sensory consequences produced by the behavior itself	Provide competing sensory input; enrich the environment noncontingently

Worked Example

Consider a functional analysis of a child's classroom disruption, run across the four test conditions plus a play control, with the target behavior recorded as responses per minute and averaged over each condition's sessions. Suppose the results are: social attention 8.4, demand (escape) 2.6, tangible 2.2, alone 1.9, and play (control) 1.5 responses per minute. Differentiation is judged against the control condition, in which noncontingent reinforcement should keep the behavior low. The attention condition runs at 8.4 ÷ 1.5 = 5.6 times the control rate, a clear elevation; the demand condition at 2.6 ÷ 1.5 = 1.73 times control is only modestly above it, and alone is essentially at control level. The analysis therefore identifies social attention as the maintaining function, and the treatment follows directly: the child is taught a simple, efficient way to recruit attention appropriately — a raised hand or a card exchanged for a moment of the teacher's time — while disruption no longer produces attention. The escape condition's slight elevation is noted as a possible secondary function to monitor, but the fivefold attention effect dominates. The arithmetic is the whole argument: without the control comparison, an 8.4-per-minute rate is just a number, but 5.6 times the rate seen when the same behavior earns nothing tells the analyst what the behavior is for.

Discussion

The strength of applied behavior analysis is also the source of its controversies: it changes behavior effectively, and the question of whose behavior should be changed, toward what goals, is not one the technology answers for itself. The seven dimensions require that behavior be socially important and that change be effective, but socially important and effective are judgments made by people, and Wolf's social-validity argument was an early acknowledgment that those judgments cannot be left to the analyst alone (Wolf, 1978). A functional analysis can tell a practitioner with precision how to reduce a behavior; it cannot tell them whether that behavior should be reduced. This is why the field's most serious internal debates are about ends, not means.

The evidence for ABA's effectiveness in teaching skills and reducing dangerous behavior is among the strongest in applied psychology for any behavioral technology, resting on thousands of replicated single-case demonstrations and, increasingly, on group trials and meta-analyses in its most studied application, autism intervention (Makrygianni et al., 2018). Yet single-case designs, powerful for demonstrating control within an individual, do not by themselves establish how well an effect generalizes across a population, and the standards for grading that evidence are exacting (Slocum et al., 2014). The result is a discipline whose within-subject rigor is unusually high and whose between-subject, population-level claims remain a live methodological question — a tension the field is aware of and actively working.

Current Directions

The most active contemporary debate concerns early intensive behavioral intervention (EIBI) for autistic children, the tradition that grew from Lovaas's landmark 1987 outcome study reporting large gains after some forty hours per week of one-to-one behavioral treatment (Lovaas, 1987). Meta-analysis supports moderate-to-large gains in adaptive behavior, language, and measured IQ from behavioral intervention, with effect sizes that tend to increase with intervention intensity and duration (Makrygianni et al., 2018). But a Cochrane review of the controlled evidence is more cautious, rating the certainty of that evidence as low to moderate and noting the scarcity of randomized trials (Reichow et al., 2018). A synthesis of four decades of the intervention literature documents a clear shift away from the adult-directed, discrete-trial format of the early programs toward more naturalistic, child-led, and developmentally framed procedures (Leaf et al., 2022).

Running alongside the effectiveness question is a sharper ethical one, raised most forcefully from within the autistic and neurodiversity communities: whether high-intensity, compliance-oriented ABA can itself be harmful, and whether normalization is a legitimate goal at all (Sandoval-Norton & Shkedy, 2019). The critique has met vigorous disagreement over its evidence and its characterization of current practice, but it has already reshaped the field's mainstream, accelerating the move toward assent-based, function-focused, and quality-of-life-oriented approaches. The clearest example is the practical functional assessment and skill-based treatment model developed by Gregory Hanley and colleagues, which builds treatment around synthesized reinforcement contingencies and the systematic teaching of communication and tolerance skills, with the child's assent as an explicit design constraint (Hanley et al., 2014).

Common Misconceptions

Negative reinforcement is a kind of punishment.: Negative reinforcement strengthens behavior by removing an aversive stimulus; punishment weakens it. The confusion is lexical: negative here means subtracted, not bad. Both reinforcement types raise response rate — positive by adding a stimulus, negative by taking one away (Cooper et al., 2020).
Applied behavior analysis is a treatment for autism.: ABA is a general applied science of behavior, not an autism treatment. Its principles are used in education, organizational management, health and safety, gerontology, and animal training; autism intervention is its most publicized application, not its definition (Baer et al., 1968).
ABA means rewarding good behavior with treats.: Reinforcement is defined by its effect on future behavior, not by being a treat, and a reinforcer for one person is neutral or aversive for another. Modern practice begins with a functional analysis of why a behavior occurs, not with a preselected reward (Iwata et al., 1994).

Glossary

Antecedent.: A stimulus or event that precedes a behavior and sets the occasion on which a consequence is available; the first term of the three-term contingency.
Discrete trial training.: A structured teaching format that breaks a skill into small steps, each taught in a distinct antecedent–behavior–consequence trial with immediate reinforcement.
Extinction.: The withholding of the reinforcer that has maintained a behavior, resulting in a gradual decline in its rate, often preceded by a temporary increase.
Functional analysis.: An experimental procedure that isolates the environmental function maintaining a behavior by systematically presenting and withholding candidate consequences.
Generality.: The extent to which a behavior change endures over time and transfers across settings, people, and related behaviors after the intervention ends.
Multiple-baseline design.: A single-case design that staggers the introduction of an intervention across behaviors, settings, or participants to show each changes only when the intervention reaches it.
Negative reinforcement.: Strengthening of a behavior by the removal of an aversive stimulus contingent on the response; negative denotes subtraction, not an undesirable outcome.
Operant conditioning.: Learning in which the future probability of a behavior is modified by its consequences; the basic science ABA applies.
Positive reinforcement.: Strengthening of a behavior by the addition of a stimulus contingent on the response, where that addition raises the behavior's future rate.
Punishment.: A consequence that decreases the future probability of the behavior it follows, by adding an aversive stimulus or removing a preferred one.
Reinforcement.: Any consequence that increases the future probability of the behavior it follows; defined functionally by its effect, not by its physical form.
Reversal design.: A single-case design (ABAB) that introduces, withdraws, and reinstates an intervention to show behavior tracks the manipulation, establishing a functional relation.
Social validity.: The judged acceptability of an intervention's goals, procedures, and outcomes to the people it affects, proposed by Wolf as a necessary complement to objective measurement.
Three-term contingency.: The antecedent–behavior–consequence unit that is the basic unit of analysis in behavior analysis.
Token economy.: A reinforcement system in which generalized conditioned reinforcers (tokens) are earned for target behaviors and later exchanged for backup reinforcers.

Key Researchers

Donald M. Baer

(1931-2002). Professor at the University of Kansas; with Wolf and Risley he co-authored the 1968 paper that defined applied behavior analysis and its seven dimensions. Wikipedia

Gregory P. Hanley

(b. 20th c.). Behavior analyst and founder of FTF Behavioral Consulting; developed the practical functional assessment and skill-based treatment model that made assent a design constraint. Faculty Page

William L. Heward

(b. 20th c.). Professor Emeritus at The Ohio State University; co-author of the field's canonical textbook and a leading voice on effective instruction. Faculty Page

Brian A. Iwata

(1948-2023). Professor at the University of Florida; developed the experimental functional analysis of problem behavior, reshaping how treatments are selected. Wikipedia - Faculty Page

O. Ivar Lovaas

(1927-2010). Professor at UCLA; pioneered early intensive behavioral intervention for autistic children and authored the influential 1987 outcome study. Wikipedia - Wikidata

B. F. Skinner

(1904-1990). Professor at Harvard University; founder of the experimental analysis of behavior and radical behaviorism, whose operant principles ABA applies. Wikipedia - Wikidata

Timothy R. Vollmer

(b. 20th c.). Professor at the University of Florida; a leading researcher on reinforcement schedules and the treatment of severe problem behavior. Google Scholar - Faculty Page

Montrose M. Wolf

(1935-2004). Professor at the University of Kansas; co-defined applied behavior analysis and introduced the concept of social validity. Wikipedia - Wikidata

Frequently Asked Questions

What is applied behavior analysis in simple terms?

Applied behavior analysis is the use of learning principles, chiefly reinforcement and its relatives, to change behavior that matters to a person, measured directly and demonstrated through controlled experiments with that individual (Baer et al., 1968).

Is ABA only for autism?

No. ABA is a general science of behavior applied across education, safety, organizational management, health, and animal training; autism intervention is its most visible application but not its definition (Cooper et al., 2020).

What is the difference between negative reinforcement and punishment?

Negative reinforcement strengthens a behavior by removing an aversive stimulus, whereas punishment weakens a behavior; the word negative means subtracted, not undesirable (Cooper et al., 2020).

What is a functional analysis?

It is an experimental procedure that identifies the environmental function a problem behavior serves (attention, escape, tangible access, or automatic reinforcement) by systematically presenting and withholding each candidate consequence (Iwata et al., 1994).

How does ABA prove that its treatment works?

It uses single-case experimental designs in which each participant is their own control, showing behavior changes when and only when the intervention is applied, reversed, or staggered across baselines (Baer et al., 1987).

Is ABA effective for autistic children?

Meta-analysis reports moderate-to-large gains in adaptive behavior and language, while a Cochrane review rates the controlled evidence as low-to-moderate certainty and calls for more randomized trials (Reichow et al., 2018).

Why is ABA controversial?

A neurodiversity-informed critique questions whether high-intensity, compliance-oriented programs can be harmful and whether normalization is a legitimate goal, prompting a shift toward assent-based practice (Sandoval-Norton & Shkedy, 2019).

Social validity is the judged acceptability of an intervention's goals, procedures, and outcomes to the people it affects, proposed by Wolf as a necessary complement to objective behavioral measurement (Wolf, 1978).

References

Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1(1), 91-97. https://doi.org/10.1901/jaba.1968.1-91

Baer, D. M., Wolf, M. M., & Risley, T. R. (1987). Some still-current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 20(4), 313-327. https://doi.org/10.1901/jaba.1987.20-313

Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.). Pearson.

Hanley, G. P., Jin, C. S., Vanselow, N. R., & Hanratty, L. A. (2014). Producing meaningful improvements in problem behavior of children with autism via synthesized analyses and treatments. Journal of Applied Behavior Analysis, 47(1), 16-36. https://doi.org/10.1002/jaba.106

Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1994). Toward a functional analysis of self-injury. Journal of Applied Behavior Analysis, 27(2), 197-209. https://doi.org/10.1901/jaba.1994.27-197

Leaf, J. B., Cihon, J. H., Ferguson, J. L., Milne, C. M., Leaf, R., & McEachin, J. (2022). Advances in our understanding of behavioral intervention: 1980 to 2020 for individuals diagnosed with autism spectrum disorder. Journal of Autism and Developmental Disorders, 52(10), 4395-4410. https://doi.org/10.1007/s10803-021-05137-y

Lovaas, O. I. (1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. Journal of Consulting and Clinical Psychology, 55(1), 3-9. https://doi.org/10.1037/0022-006X.55.1.3

Makrygianni, M. K., Gena, A., Katoudi, S., & Galanis, P. (2018). The effectiveness of applied behavior analytic interventions for children with autism spectrum disorder: A meta-analytic study. Research in Autism Spectrum Disorders, 51, 18-31. https://doi.org/10.1016/j.rasd.2018.03.006

Reichow, B., Hume, K., Barton, E. E., & Boyd, B. A. (2018). Early intensive behavioral intervention (EIBI) for young children with autism spectrum disorders (ASD). Cochrane Database of Systematic Reviews, 5, CD009260. https://doi.org/10.1002/14651858.CD009260.pub3

Sandoval-Norton, A. H., & Shkedy, G. (2019). How much compliance is too much compliance: Is long-term ABA therapy abuse? Cogent Psychology, 6(1), 1641258. https://doi.org/10.1080/23311908.2019.1641258

Skinner, B. F. (1953). Science and human behavior. The Free Press.

Skinner, B. F. (1957). Verbal behavior. Copley Publishing Group.

Ferster, C. B., & Skinner, B. F. (1957). Schedules of reinforcement. B. F. Skinner Foundation.

Slocum, T. A., Detrich, R., Wilczynski, S. M., Spencer, T. D., Lewis, T., & Wolfe, K. (2014). The evidence-based practice of applied behavior analysis. The Behavior Analyst, 37(1), 41-56. https://doi.org/10.1007/s40614-014-0005-2

Wolf, M. M. (1978). Social validity: The case for subjective measurement or how applied behavior analysis is finding its heart. Journal of Applied Behavior Analysis, 11(2), 203-214. https://doi.org/10.1901/jaba.1978.11-203

Applied Behavior Analysis