functional-behavior-assessment

# Functional Behavior Assessment (FBA) ## Purpose and Legal Context A Functional Behavior Assessment is a systematic process for identifying the environmental variables that reliably predict and maintain problem behavior. Under IDEA (2004), an FBA is mandated when a student with a disability faces disciplinary removal exceeding 10 school days or when the behavior is determined to be a manifestation of the student's disability. Beyond legal mandate, FBA is the ethical and empirical foundation for any function-based intervention. The goal is never simply to describe the topography of behavior — it is to identify the **function**: the reinforcement contingency maintaining the behavior. Without functional understanding, intervention is guesswork. ## Three Levels of Assessment ### Level 1: Indirect Assessment Indirect methods gather information through interviews and rating scales without direct observation. They are efficient for generating initial hypotheses but are insufficient as sole data sources. **Structured Rating Scales:** - **Motivation Assessment Scale (MAS; Durand & Crimmins, 1988):** 16 items rating likelihood of behavior across four functions (sensory, escape, attention, tangible). Quick but psychometric limitations — use as screening only. - **Functional Analysis Screening Tool (FAST; Iwata & DeLeon, 1996):** Checklist format identifying likely maintaining variables. Better inter-rater agreement than MAS for escape and attention functions. - **Questions About Behavioral Function (QABF; Matson & Vollmer, 1995):** 25 items across five subscales (attention, escape, nonsocial/automatic, physical, tangible). Strongest psychometric support among indirect tools. Score ≥ 4 on a subscale suggests that function. - **Functional Assessment Interview (FAI; O'Neill et al., 1997):** Comprehensive semi-structured interview covering behavior description, ecological events, setting events, antecedents, consequences, communicative function, history, and prior interventions. **Open-Ended Interviews:** Conduct with all relevant stakeholders (caregivers, teachers, direct-care staff, the individual when possible). Ask: 1. What does the behavior look like? (Operational definition) 2. When/where does it happen most? Least? 3. Who is present? What activity is occurring? 4. What happens right before? 5. What happens right after? 6. What does the individual seem to get or avoid? 7. What has been tried? What worked, even briefly? ### Level 2: Descriptive Assessment Descriptive methods involve direct observation in the natural environment. They capture actual antecedent-behavior-consequence sequences but cannot establish functional relations (correlation ≠ causation). **ABC Narrative Recording:** Observer writes a running narrative of events before, during, and after each instance of the target behavior. Rich qualitative data but difficult to quantify and subject to observer bias. **Structured ABC Data Collection:** Pre-formatted data sheets with columns for time, setting, antecedent, behavior, consequence, and perceived function. Allows quantification of conditional probabilities — how often does behavior X follow antecedent Y and precede consequence Z? **Scatterplot Analysis (Touchette, MacDonald, & Langer, 1985):** Grid mapping behavior occurrence across time intervals and days. Reveals temporal patterns — does behavior cluster during transitions, after lunch, on Mondays? Identifies setting events and slow triggers. **Conditional Probability Analysis:** Calculate: P(behavior | antecedent) vs P(behavior | no antecedent). A high conditional probability for a specific antecedent suggests (but does not prove) a functional relation. Also calculate: P(consequence | behavior) to evaluate which consequences reliably follow behavior. ### Level 3: Experimental (Functional Analysis) When indirect and descriptive data yield ambiguous or conflicting hypotheses, experimental functional analysis (see `functional-analysis` skill) is the gold standard. The FBA practitioner must decide whether available data are sufficient or whether FA is needed. ## Integrating Multiple Data Sources Triangulation across methods strengthens confidence: 1. Conduct indirect assessment (QABF + FAI or open-ended interview) 2. Conduct descriptive assessment (structured ABC + scatterplot, minimum 3-5 observation sessions) 3. Compare hypotheses across methods 4. If convergence → write function statement with confidence 5. If divergence → consider experimental FA, additional observations, or reconceptualize the behavior Weight descriptive data over indirect data when they conflict. Weight experimental data over both. ## Writing Function Statements Use the four-term contingency format: > **When** [antecedent/setting event], **the individual** [operationally defined behavior], **which results in** [consequence], **and is maintained by** [function: specific reinforcement contingency]. **Examples:** - When presented with non-preferred academic tasks (math worksheets) during independent work time, Marcus engages in head-hitting (forceful contact of closed fist to temporal region of head, 3+ times per minute), which results in task removal by the teacher, maintained by **negative reinforcement (escape from demands)**. - When attention is directed away from Amara toward a peer during free play, Amara engages in screaming (vocalizations above conversational volume lasting 3+ seconds), which results in adult verbal reprimand or redirection, maintained by **positive reinforcement (adult attention)**. ## The Four Primary Functions | Function | Reinforcement Type | Antecedent Pattern | Consequence Pattern | |----------|-------------------|--------------------|--------------------| | Attention | Positive (+SR) | Low attention, attention diverted | Social interaction delivered | | Escape/Avoidance | Negative (−SR) | Demand, non-preferred activity | Task removal/reduction | | Tangible/Activity | Positive (+SR) | Restricted access, item/activity removed | Item/activity delivered | | Automatic (sensory) | Positive or Negative | Variable; may not correlate with social antecedents | Sensory stimulation produced or aversive stimulation attenuated | ### Automatic Reinforcement — Positive vs Negative - **Automatic positive reinforcement:** Behavior produces sensory stimulation (e.g., hand-flapping producing visual input, vocal stereotypy producing auditory input). Behavior increases in enriched AND impoverished environments; may increase during alone/low-stimulation conditions. - **Automatic negative reinforcement:** Behavior attenuates aversive sensory input (e.g., head-pressing to relieve pain, SIB co-occurring with otitis media). Suspect when behavior correlates with illness, biological events, or aversive environmental stimuli (noise, light). ## FBA vs FA Decision Framework **FBA alone is sufficient when:** - Indirect and descriptive data converge on a single function - The behavior is not dangerous or high-intensity - Intervention can be implemented and monitored quickly - The setting does not permit experimental manipulation **FA is needed when:** - Data sources yield conflicting hypotheses - Multiple functions are suspected and prioritization is required - Previous function-based interventions have failed - The behavior is severe/dangerous and precise function identification is critical - Automatic reinforcement is hypothesized (difficult to confirm without experimental control) ## FBA Report Structure 1. **Referral Information:** Reason for assessment, referral source, relevant history 2. **Background/Records Review:** Diagnoses, medications, educational placement, prior assessments 3. **Behavior Description:** Operational definitions of all target behaviors with examples and non-examples 4. **Assessment Methods:** Tools used, dates, informants, observation settings and duration 5. **Indirect Assessment Results:** Summary of interviews and rating scales 6. **Descriptive Assessment Results:** ABC data summaries, scatterplot patterns, conditional probabilities 7. **Data Integration and Hypothesis:** Convergence/divergence across data sources 8. **Function Statement(s):** One per target behavior 9. **Recommendations:** Function-based intervention components, replacement behaviors, antecedent modifications 10. **Appendices:** Data sheets, graphs, completed rating scales ## Quality Checklist - [ ] Behavior operationally defined (observable, measurable, with examples) - [ ] Multiple informants interviewed - [ ] Standardized indirect tool administered (QABF, FAST, or MAS) - [ ] Minimum 3-5 direct observation sessions across settings/times - [ ] Setting events and slow triggers considered - [ ] Conditional probabilities calculated from ABC data - [ ] Data sources triangulated - [ ] Function statement follows four-term format - [ ] Recommendations are function-matched - [ ] Report distinguishes between hypothesized and confirmed functions ## Key References - Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1982/1994). Toward a functional analysis of self-injury. *Journal of Applied Behavior Analysis*, 27, 197–209. - Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003). Functional analysis of problem behavior: A review. *Journal of Applied Behavior Analysis*, 36, 147–185. - O'Neill, R. E., Albin, R. W., Storey, K., Horner, R. H., & Sprague, J. R. (2015). *Functional Assessment and Program Development for Problem Behavior* (3rd ed.). Cengage. - Matson, J. L., & Vollmer, T. R. (1995). User's guide: Questions About Behavioral Function (QABF). Scientific Publishers. - Touchette, P. E., MacDonald, R. F., & Langer, S. N. (1985). A scatter plot for identifying stimulus control of problem behavior. *Journal of Applied Behavior Analysis*, 18, 343–351.