Comparative, Component, and Parametric Analyses
Key Takeaways
- Comparative analyses ask which intervention or condition produces better effects.
- Component analyses ask which part of a treatment package is necessary or active.
- Parametric analyses ask which value, amount, intensity, or schedule parameter is most effective.
- These analyses should still preserve single-case logic through repeated measurement and planned comparisons.
Analyses that refine intervention decisions
A comparative analysis evaluates relative effects of two or more conditions. In practice, this may mean comparing noncontingent reinforcement with differential reinforcement, or comparing two prompting strategies. Multielement and reversal formats are common, but the design must still control carryover and discrimination problems.
A component analysis identifies which parts of a treatment package are necessary. If a package includes antecedent prompts, response blocking, and reinforcement, the analyst may remove or add components systematically. The goal is not novelty; it is finding the simplest effective package.
A parametric analysis evaluates different values of the same variable. Examples include schedule thinning values, token exchange ratios, prompt delays, session durations, or reinforcer magnitudes. The question is not whether reinforcement works, but which parameter value produces acceptable outcomes.
| Analysis | Question | Example |
|---|---|---|
| Comparative | Which condition works better? | DRA versus NCR |
| Component | Which element is necessary? | Full package versus package without prompts |
| Parametric | Which value works best? | FR 1, FR 3, and FR 5 schedules |
These analyses can improve social validity and efficiency. A treatment package may be effective but too effortful for routine use. A component analysis can remove unnecessary pieces, while a parametric analysis can identify a leaner schedule that still maintains behavior.
Exam cue
If the options differ by kind, think comparative. If the options remove or add parts of a package, think component. If the options change the amount or value of one variable, think parametric.
A BCBA alternates two interventions to determine whether differential reinforcement or response cost produces lower rates of disruption. What type of analysis is this?
An intervention package includes prompts, token reinforcement, and escape extinction. The BCBA systematically removes one element at a time to see which pieces are needed. What analysis is this?
A BCBA tests whether a 10-second, 20-second, or 40-second prompt delay produces the best balance of independence and errors. What type of analysis is this?