BCBA Exam (Board Certified Behavior Analyst) — All Questions

In a study of whether a token system increases homework completion, the token system is the

• a.baseline condition
• b.dependent variable
• c.independent variable✓
• d.confounding variable

The independent variable is what the experimenter manipulates to see its effect, here the token system. Homework completion, the measure that may change, is the dependent variable. A confounding variable is an uncontrolled factor that could explain the result, and baseline is the condition without the token system.

A reading intervention produced a clear effect for three second-graders in one school. A reviewer asks whether it would work for older students in other schools. This question concerns

• a.treatment integrity
• b.external validity✓
• c.interobserver agreement
• d.internal validity

External validity is the extent to which a functional relation holds across other participants, settings and behaviors, which is what the reviewer asks. Internal validity concerns whether the independent variable, rather than something else, produced the change in the original study. Interobserver agreement and treatment integrity concern measurement of behavior and implementation, not generality.

A toddler's spoken vocabulary grows steadily over the six months of a language intervention. Which threat to internal validity most plausibly competes with the intervention as an explanation?

• a.Regression to the mean
• b.Diffusion of treatment
• c.Instrumentation
• d.Maturation✓

Maturation refers to changes produced by developmental processes over time; toddlers' vocabulary grows naturally over six months, so it could account for the gains. Instrumentation concerns changes in how measurement is done, regression to the mean concerns extreme initial scores drifting back, and diffusion of treatment involves a comparison group receiving the treatment; none of these follows from the scenario as described.

In an A-B-A-B reversal design, behavior returns toward baseline levels when the intervention is withdrawn in the second A phase. This return provides

• a.prediction
• b.verification✓
• c.replication
• d.generalization

Baseline logic in single-case designs has three elements: prediction from stable baseline data, verification that behavior would have continued at baseline levels had the intervention not been introduced (shown by the return in the second A phase), and replication of the effect when the intervention is reintroduced in the second B phase. Generalization is not an element of baseline logic.

A BCBA wants to evaluate a program teaching three different social skills that are unlikely to be lost once learned. Withdrawing the program would not be expected to reverse the behavior. Which design is most appropriate?

• a.Changing-criterion design
• b.A-B-A-B reversal design
• c.B-A-B withdrawal design
• d.Multiple-baseline design✓

When target behaviors are not expected to reverse, a multiple-baseline design demonstrates control by staggering the intervention across behaviors (or settings or people) and showing change only when each is treated, as Baer, Wolf and Risley described. Reversal and B-A-B designs depend on behavior returning toward baseline when treatment is withdrawn. A changing-criterion design evaluates stepwise changes in a single behavior.

A smoker's daily cigarette limit is lowered in steps (15, then 12, then 9), and smoking closely tracks each new limit before the next step. Which design is this?

• a.Changing-criterion✓
• b.Reversal
• c.Multielement
• d.Multiple baseline across settings

Hartmann and Hall (1976) described the changing-criterion design: a single behavior is required to meet a series of stepwise criteria, and control is shown when behavior changes to match each new criterion. A multiple baseline staggers treatment across tiers, a multielement design rapidly alternates conditions, and a reversal design withdraws treatment.

To compare two prompting methods for the same learner, a BCBA alternates them across sessions in a randomized order and graphs each method as its own data path. This is

• a.a multielement design✓
• b.an A-B design
• c.a multiple-baseline design
• d.a changing-criterion design

Rapidly alternating two or more conditions and comparing their separate data paths is a multielement (alternating-treatments) design, well suited to comparing interventions within one person. A changing-criterion design sets stepwise performance criteria. A multiple baseline staggers one intervention across tiers. An A-B design has a single baseline and a single intervention phase and cannot compare two methods.

A researcher delivers 1, 2, or 4 tokens per correct response in different phases to find how reinforcer magnitude affects a student's work rate. This is

• a.a component analysis
• b.a functional analysis
• c.a comparative analysis
• d.a parametric analysis✓

A parametric analysis examines the effects of different values of one independent variable, here reinforcer magnitude. A component analysis isolates which parts of a treatment package are responsible for its effect, a comparative analysis compares two or more different treatments, and a functional analysis identifies the reinforcers maintaining problem behavior.

In a multiple-baseline design across three classrooms, disruptive behavior in the second classroom decreases sharply when the intervention is introduced in the first classroom, before the second classroom receives it. What does this pattern mean for the analysis?

• a.It confirms the intervention's effect in the second tier
• b.It weakens the demonstration of experimental control✓
• c.It shows ideal verification of the first tier
• d.It shows that the tiers were fully independent

Multiple-baseline logic requires untreated tiers to stay stable while an earlier tier changes; that stability verifies that change occurs only when the intervention is applied. Change in an untreated tier suggests the tiers are not independent, or that another variable is at work, so experimental control is weakened. It does not confirm an effect or show independence in the tiers.