Behavior Assessment
Three tiers of functional behavior assessment
- Indirect - interviews, rating scales (FAST, MAS, QABF), records review. Fast and low effort, low confidence.
- Descriptive - ABC recording, scatterplots, conditional probability analysis. Correlational: tells you when and with whom, not why.
- Functional analysis - systematic manipulation of antecedents and consequences (attention, escape/demand, tangible, alone/ignore, and a play control). Only this tier demonstrates a function.
Preference and reinforcer assessment
- Preference assessments: paired choice (more trials, more confidence), MSWO (rank order), free operant (low effort, good when forced choice is poorly tolerated), single stimulus.
- Reinforcer assessments verify that a preferred item actually reinforces behavior. Preference is not reinforcement.
Other tools
Scatterplots reveal time-of-day patterns; conditional probability shows which consequence most often follows the behavior. Standardized instruments (Vineland, ABLLS, AFLS, VB-MAPP) inform skill and curriculum decisions, not function.
Traps
- Concluding function from a descriptive assessment. "Behavior occurred most often during math" is a hypothesis, not an escape function.
- Confusing preference assessment results with treatment effectiveness.
- A functional analysis with no control condition, or a control contaminated by attention for problem behavior.
- Missing automatic reinforcement when behavior persists in the alone condition.
Worked example
Aggression occurs in 80% of demand intervals and 10% of free play, and staff deliver a break after 90% of aggression. That supports an escape hypothesis; confirm it with a functional analysis containing escape (demand, break contingent), attention, tangible, and play conditions. If aggression is elevated only in escape, the function is negative reinforcement, and treatment should teach a functionally equivalent communication response.