BCBA cheat sheet

Radical behaviorism treats private events such as thoughts and feelings as behavior, explained by the same environmental variables that explain public behavior.

Methodological behaviorism excludes private events from analysis; radical behaviorism includes them, but as dependent variables rather than causes of behavior.

Selectionism operates at three levels: phylogeny (natural selection), ontogeny (operant learning), and cultural selection of practices transmitted across generations.

Determinism holds that behavior is lawful and caused by environmental and historical variables; behavior analysis rejects free-will or agentic explanations.

Empiricism requires objective observation and replication; appeals to authority, tradition, or consensus are not accepted as evidence.

Parsimony: when two accounts have equal explanatory power, the one requiring fewer assumptions is preferred.

Pragmatism: the meaning and truth of a concept rest on its observable effects and practical utility, not on correspondence to an inner state.

Explanatory fictions are circular labels treated as causes (for example, 'she cried because she was sad') and are rejected as explanations.

Mentalism explains behavior by hypothetical inner agents (mind, traits, cognitions); behavior analysis instead seeks the environmental variables controlling behavior.

Positive reinforcement: a stimulus is added contingent on behavior and behavior increases. Negative reinforcement: a stimulus is removed contingent on behavior and behavior increases.

Positive punishment: a stimulus is added and behavior decreases. Negative punishment: a stimulus is removed and behavior decreases.

Reinforcement and punishment are defined by their effect on future response frequency, not by whether the event appears pleasant or aversive.

An SD is correlated with availability of reinforcement for a response; an S-delta is correlated with extinction for that response.

Establishing operations increase reinforcer value and evoke the relevant behavior; abolishing operations decrease reinforcer value and suppress it.

Motivating operations have two effects: value-altering (reinforcer efficacy) and behavior-altering (momentary frequency of the relevant response class).

Extinction withholds the reinforcer maintaining behavior; typical effects include extinction burst, resurgence, and spontaneous recovery.

Shaping arranges differential reinforcement of successive approximations; chaining links component responses identified through a task analysis.

Response classes are defined functionally: topographically different responses producing the same consequence belong to one operant class.

Rate equals total responses divided by observation time; rate is the preferred measure for free-operant behavior.

Percentage equals (occurrences ÷ opportunities) × 100 and is used when trials or opportunities are discrete and countable.

Total-count IOA: smaller count ÷ larger count × 100. Mean count-per-interval IOA averages agreement across intervals.

Trial-by-trial IOA: trials with exact agreement ÷ total trials × 100.

Acceptable IOA is generally 80% or higher, with 90% or higher preferred for stringent or publication standards.

Partial interval recording tends to overestimate behavior; whole-interval recording tends to underestimate it; momentary time sampling is generally most accurate at short intervals.

On a cumulative graph the slope equals response rate; a flat line indicates no responding and a downward slope indicates a decreasing rate.

Visual analysis evaluates level, trend, variability, immediacy of change, overlap between conditions, and consistency across similar phases.

Latency is the time from stimulus onset to response onset; interresponse time is the time between successive responses.

Reversal (A-B-A-B) designs demonstrate control when behavior changes with each introduction and withdrawal of the independent variable; they require reversible behavior.

Multiple baseline designs stagger intervention onset across behaviors, settings, or participants and do not require withdrawal of treatment.

Alternating treatments (multielement) designs compare conditions rapidly; sequence effects are controlled by randomization or counterbalancing.

Changing criterion designs verify effects through stepwise shifts in the criterion required for reinforcement.

Internal validity threats include maturation, history, testing, instrumentation, regression to the mean, and attrition.

Multiple treatment interference is a specific risk in multielement designs when the effects of one condition carry over to another.

Parametric analysis varies the value of one independent variable; component (dismantling) analysis removes elements of a treatment package.

Prediction, verification, and replication are the three elements of baseline logic in single-case experimental designs.

The BACB Ethics Code for Behavior Analysts took effect January 1, 2022, with four core principles: benefit others; treat others with compassion, dignity, and respect; behave with integrity; ensure competence.

Informed consent must disclose procedures, risks, benefits, and alternatives, note the right to withdraw, and is an ongoing process rather than a single event.

Confidentiality may be breached for mandated reporting, court orders, or imminent danger to self or others, and only information necessary to the purpose is disclosed.

Assent is obtained from clients who cannot legally consent, in understandable terms; continued dissent is honored when feasible.

Least restrictive, reinforcement-based procedures are preferred; punishment-based procedures require justification, increased oversight, and ongoing data.

Behavior analysts practice within their demonstrated competence and refer when a case falls outside their scope.

Records including data, consent documentation, and supervision documents are retained for at least seven years under BACB requirements.

Multiple relationships are avoided; when unavoidable they are documented and steps are taken to mitigate risk of harm or impaired objectivity.

Supervisors must observe at least 5% of a trainee's experience hours and hold at least one individual supervision meeting per month.

FBA methods fall into three categories: indirect (interviews, rating scales), descriptive (ABC recording, scatterplots), and functional analysis (experimental manipulation).

Standard functional analysis conditions are alone, attention, demand/escape, tangible, and play (control).

Indirect measures such as the FAST, MAS, and QABF generate hypotheses but cannot demonstrate functional relations.

Descriptive assessments identify correlations between environmental events and behavior, not causal relations.

Preference assessments include single stimulus, paired stimulus, multiple stimulus with replacement, and MSWO (selected item removed after each trial).

MSWO produces a preference ranking more quickly than paired choice but may obscure the strength of preference for the first-selected item.

Reinforcer assessments (concurrent operants, progressive ratio) verify that a preferred stimulus actually functions as reinforcement.

Scatterplots display behavior by time of day or activity to reveal temporal patterns useful for hypothesis generation.

Functional analysis of severe behavior requires trained implementers, safety protocols, and controlled conditions with safeguards.

DRA reinforces an alternative response while the reinforcer maintaining problem behavior is withheld.

DRO delivers reinforcement after an interval containing no target behavior; a resetting DRO restarts the interval after each occurrence.

DRI reinforces a response topographically incompatible with the target; DRL and DRH change the required rate of responding.

Functional communication training is DRA in which the alternative response is a socially acceptable communication response matched to the functional analysis.

Noncontingent reinforcement delivers the reinforcer on a time-based schedule independent of behavior, reducing responding partly through an abolishing operation.

Extinction is most effective when combined with reinforcement of an alternative response; side effects include extinction bursts and resurgence.

Prompting strategies include least-to-most prompting, most-to-least prompting, progressive time delay, and stimulus fading.

Antecedent interventions include high-probability request sequences (behavioral momentum), choice, priming, and task modification.

Response cost removes a specified amount of a reinforcer contingent on behavior; time-out requires a period with no access to reinforcement.

Interventions are selected to match the function identified by the FBA; function-matched treatments generally outperform arbitrary or default treatments.

Evidence-based selection weighs research support alongside client and caregiver values, context, cost, and feasibility.

Social validity is assessed for goals, procedures, and outcomes, often through stakeholder rating scales.

Treatment integrity data come from direct observation or permanent products; low integrity predicts poor outcomes independent of the intervention's design.

Generalization tactics include teaching multiple exemplars, training loosely, programming common stimuli, and using indiscriminable contingencies.

Maintenance is promoted by thinning reinforcement schedules, fading prompts, and transferring control to natural contingencies.

Progress monitoring compares intervention data to baseline using preset decision rules, such as three consecutive data points at or below criterion.

Component analysis identifies which elements of a treatment package are necessary before large-scale implementation.

BACB supervision requires at least 5% of the trainee's experience hours, at least one individual supervision contact per month, and documented direct observations.

Behavioral skills training consists of instruction, modeling, rehearsal, and feedback, repeated until the trainee meets a performance criterion.

Performance feedback is most effective when immediate, specific, and delivered with both positive and corrective components.

Pyramidal training has experts train a small group who then train others, extending supervisory reach while fidelity is checked.

Supervision contracts specify roles, expectations, evaluation methods, and conditions for terminating supervision.

Staff performance is measured by direct observation, permanent products, and self-monitoring rather than self-report alone.

Tasks are delegated only after verifying the supervisee's competence and confirming the activity falls within legal and ethical scope.

Supervision documentation includes signed agreements, observation records, and monthly verification forms.

Antecedent staff-management tactics include task clarification, goal setting, and self-monitoring interventions.