Every feature in Behavior Celeration is grounded in peer-reviewed research from applied behavior analysis, precision teaching, and learning science. Here's how each piece works — and why.
Originated by B.F. Skinner (1968) and refined through errorless learning research by Terrace (1963), prompt fading teaches concepts in small, manageable steps. Learners progress through four tiers:
Complete answer provided. Learner reads and absorbs. Builds familiarity without pressure.
Key terms blanked. Learner fills in missing pieces using context clues and emerging knowledge.
First-word hint only. Learner must recall the full answer from a single cue — building fluency.
No prompt. Full recall. The test of true mastery.
Skinner 1968 Terrace 1963 Touchette & Howard 1984
Developed by Ogden Lindsley (1971) from Skinner's cumulative recorder, the Standard Celeration Chart (SCC) uses a semi-logarithmic scale to visualize learning speed — the "celeration" rate.
Unlike linear charts that flatten learning gains, the SCC's proportional vertical axis (1, 10, 100, 1000) reveals true multiplicative growth. A doubling of correct responses looks the same whether you're at 10 or 1000.
Our app plots your daily performance on the SCC so you can see your own learning celeration — not just raw scores. The slope of your celeration line tells you exactly how fast you're learning.
Lindsley 1971 White & Haring 1980 Kubina & Yurich 2012
Accuracy isn't enough. C. Warren Binder (1996) demonstrated that fluency — responding both accurately AND quickly — produces four critical outcomes:
The skill is performed correctly.
The skill persists over time without re-teaching.
The skill can be performed for extended periods.
The skill transfers to new, untrained contexts.
SAFMEDS (Say All Fast Minute Every Day Shuffled), introduced by Elaine Haughton (1980), is our 60-second sprint format. You flip through as many frames as possible in one minute, building automatic recall — not just recognition.
Binder 1996 Haughton 1980 Kubina & Morrison 2000
Spaced repetition optimizes review timing by scheduling items just before you'd forget them. Ebbinghaus (1885) discovered the forgetting curve; modern algorithms like FSRS (Settles & Meeder, 2016) compute optimal intervals mathematically.
Our FSRS-ABA hybrid adapts the standard FSRS curve for ABA-specific content:
The algorithm uses the FSRS stability formula R=(1+19t/S)^-1 where R = retrieval
probability, t = time since last review, and S = stability. Your individual performance
continuously recalibrates the model.
Ebbinghaus 1885 Pavlik & Anderson 2008 FSRS 2016
Our answer matching uses all-target-token validation with typo tolerance. A response is correct only if it contains ALL key terms from the target answer — no partial matches, no concept flips.
This prevents false positives like "negative reinforcement" matching "positive reinforcement" (which early prototype versions caught and fixed). The system also handles minor spelling variations while rejecting fundamentally wrong answers.
BACB 6th Ed TCO Internal R&D
All content maps directly to official BACB test content outlines:
Frames are hand-authored by behavior analysts for accuracy-critical content, then supplemented with AI-generated frames (marked "unaudited") that cover every task in the list. All content requires BCBA review before certification claims can be made.
BACB Official Quality Assurance