Variables & Validity
This type of variable is manipulated or controlled by the researcher to examine its effect on behavior.
What is an independent variable?
This defining feature of single-case experimental designs means the participant's own baseline data serves as the comparison for intervention-phase data.
What is the individual serving as their own control?
In this design, behavior is expected to improve in Phase B, return toward baseline levels in the second Phase A, and improve again in the second Phase B.
What is a reversal (or ABAB) design?
A behavioral safety program includes written checklists, supervisor feedback, and group-contingent reinforcement. The program works, but the team does not know which element is responsible. This type of analysis is needed.
What is a component analysis?
A BCBA wants to evaluate whether performance feedback increases staff protocol fidelity. Staff fidelity is likely to drop when feedback is removed, and withdrawal raises no ethical concerns. This is the most appropriate design
What is a reversal (ABAB) design?
A functional relationship, an operational definition, and repeated measurement are all characteristics of high quality in this type of validity.
What is internal validity?
This defining feature is what separates a single-case experiment from a case study — it is the mechanism through which alternative explanations are made implausible.
What is replication?
This design type is selected when the target behavior is a skill that, once acquired, is unlikely to reverse when the intervention is withdrawn.
What is a multiple baseline design?
A contingency management program uses FR1 reinforcement. The team wants to know whether FR3 or FR5 would maintain outcomes while reducing reinforcer delivery burden. This type of analysis is needed.
What is a parametric analysis?
A BCBA is teaching functional communication training to replace self-injurious behavior. The team has decided that withdrawing FCT after it is effective is not ethically acceptable. This design achieves replication without requiring withdrawal.
What is a multiple baseline design?
When a researcher introduces a token economy and a new visual schedule at the same time, this threat to internal validity is present.
What is concurrent treatment (or confounding)?
A stable, countertherapeutic baseline is required for this defining feature because it establishes what the DV would look like if the intervention were never introduced.
What is prediction?
A BCBA wants to compare whether errorless learning or error correction produces faster acquisition of sight words for a student with dyslexia. Both conditions will be applied on alternating days. This is the design type.
What is a multielement (or alternating treatments) design?
Multiple treatment interference, where exposure to one condition affects performance in another, is the primary threat to validity for this type of analysis.
What is a comparative analysis?
A supervisor reviews a proposed study and notes that the baseline for Tier 2 was only two data points before the intervention was introduced. This is the design quality problem.
What is insufficient baseline length (or inadequate baseline stability)?
A client was selected for intervention because she had the highest rate of aggression in the facility that month. When her rates improve the following month, this threat is the most parsimonious explanation.
What is regression to the mean?
This is the strength of single-case designs that makes them the more appropriate choice when the question is 'did this IV change this behavior for this person?' rather than 'does this intervention work on average?'
What is sensitivity to individual differences (or the individual as the unit of analysis)?
A behavioral health coach implements a reinforcement program to increase daily step counts, beginning at 3,000 steps and raising the requirement by 1,500 steps every two weeks. This is the design type.
What is a changing criterion design?
A subtractive design begins with the full intervention package and systematically removes components one at a time. This is one approach to conducting this type of analysis.
What is a component analysis?
A BCBA is conducting a multiple baseline across three school settings for a student with ADHD, but discovers that self-monitoring skills introduced in Setting 1 are already being used spontaneously in Settings 2 and 3. This is the design threat that has occurred.
What is generalization across tiers?
A researcher reports that 85% of program completers improved — but 40 of the original 60 participants dropped out before completion, mostly because they were not responding to treatment. This threat explains why the reported outcome is misleading.
What is attrition (or differential dropout)?
A single-case study with 40 participants across multiple clinics and diagnostic groups has this type of validity, the degree to which findings generalize beyond the study's specific conditions.
What is external validity?
In the Nelson and Cone (1979) token economy study, the token program was introduced to personal hygiene, personal management, ward work, and social skills in sequential fashion. This was which type of multiple baseline?
What is a multiple baseline across behaviors?
A researcher establishes that a behavioral intervention works, then systematically varies the delay to reinforcement from immediate to 5 seconds to 30 seconds to identify the optimal value. This is this type of analysis.
What is a parametric analysis?
Before a single-case study begins, the researcher specifies that if the participant's performance drops below 60% accuracy in any subthreshold condition for three consecutive sessions, that condition will be discontinued. This pre-specified criterion is called this.
What is a stopping rule?