Baseline Logic Slides Baseline Logic and Prediction
The Foundation of Data-Driven Adjustment
Lesson 01
The Lucky Streak
Imagine a student who usually fails math quizzes suddenly gets an 85%.
Option A
"They finally understand the material! I'll move them to the advanced group today."
Option B
"I need to see three more quizzes before I believe this isn't just a fluke."
Why is "Wait and See" the clinical choice?
The Three Pillars of Baseline Logic
01
Prediction
The anticipated outcome of a future measurement if the current conditions remain unchanged.
02
Verification
Confirming that the behavior would have stayed the same if the intervention hadn't happened.
03
Replication
Repeating the observation to show the intervention consistently changes the behavior.
Determining Stability
Criterion Check
Variability
The frequency and degree to which multiple measures of behavior yield different outcomes.
Trend
The overall direction taken by a data path (Improving, Worsening, or Stable).
Level
The value on the vertical axis around which a series of data points converges.
A stable baseline has minimal variability and a zero/neutral trend.
The Clinical Decision
Go!
Stable baseline with low variability OR behavior is dangerous and requires immediate intervention regardless of stability.
Wait...
Variable baseline or trend is already moving in the desired direction (the behavior is "fixing itself").
Stop & Recalibrate
Baseline data shows that the measurement system itself is flawed or inconsistent.
Baseline Logic Teacher Guide Baseline Logic Facilitator Guide
Lesson 1: Prediction, Verification, and Replication
Executive Summary
This lesson establishes the scientific foundation of single-subject design. Students will transition from viewing data as "just numbers" to viewing it as a predictive tool. By the end of this session, students should be able to justify why an intervention should or should not begin based on the stability of baseline data.
Key Objectives
Define Prediction, Verification, and Replication.
Assess data stability using trend, level, and variability.
Make clinical decisions on intervention timing.
Hook Facilitation: The Lucky Streak
Setup: Present the scenario of a student who consistently scores 40-50% on math quizzes suddenly scoring an 85% once. Ask the class: "Is the problem solved?"
The "Intuitive" Response
They did it! Move them up! (Danger: False Positive/Fluke)
The "Data-Driven" Response
We need more data to predict if this 85% is a new trend or an outlier.
Analogy: Think of baseline like a weather forecast. One sunny day doesn't mean winter is over.
Instructional Talking Points
1
Prediction: The "What If" Phase
Emphasize that we collect baseline not just for history, but to project the future. If we don't intervene, where is this behavior heading? Without a clear prediction, we cannot measure the effect of our intervention.
2
Stability: The Clinical "Green Light"
Explain that "Stable" doesn't mean a straight line; it means predictable . A steady upward trend is stable because we know where it will be tomorrow. A jagged, unpredictable zig-zag is variable and makes intervention effects hard to prove.
3
Ethical Exceptions
Critical Note: If a behavior is dangerous (e.g., self-injury, aggression), we do NOT wait for a stable baseline. Ethical safety overrides the need for perfect data logic.
Discussion Prompts
"What happens if we intervene while a behavior is already improving on its own?"
Answer: We commit a 'False Positive'—we think our intervention worked, but the behavior was already changing.
"Why do we need at least 3-5 data points for a baseline?"
Answer: Two points make a line; three points start to show a trend; five points provide confidence in stability.
End of Facilitator Guide — Lesson 1
Baseline Prediction Worksheet Predicting the Path
Applied Baseline Logic Practice
Name:
Date:
1
Evaluating Stability
Analyze the baseline data below. For each scenario, determine if the data is Stable or Variable and provide your clinical recommendation: Intervene or Wait.
Graph A: Target Behavior Frequency
Status:
Recommendation:
Justification:
Graph B: Target Behavior Frequency
Status:
Recommendation:
Justification:
2
Drawing the Future
Task: Below is a baseline for a behavior we want to decrease (e.g., Off-task verbalizations). 1. Draw a dashed line (---) to predict what the data would do if no intervention occurs. 2. Draw a solid line (—) to show what the data should look like if a successful intervention begins at Session 6.
Frequency
Sessions
12345 678910
Intervention Phase
3
Reflective Analysis
If your baseline data was highly variable, but the behavior was "Self-Injurious Head Banging," would you wait for stability? Explain your ethical rationale.
Treatment Fidelity Slides Treatment Fidelity
Did We Actually Do the Intervention?
Lesson 02
The Failed Cookie Mystery
A baker tries a new cookie recipe. The cookies come out flat and salty.
Initial Reaction:
"This recipe is terrible! Throw it away and find a new one."
But Wait! Investigation reveals:
The baker forgot the baking soda.
The oven was 50 degrees too cold.
They used salt instead of sugar.
Is it a Recipe Failure or a Fidelity Failure?
What is Treatment Fidelity?
"The extent to which an intervention is implemented as intended by the researchers or practitioners."
Procedural
Following the steps of the plan.
Dosage
Implementing for the right amount of time.
Quality
Delivering the plan with appropriate skill.
Why Fidelity Fails
Trainer Variables
Lack of training or clear protocols.
High complexity of the intervention.
Lack of ongoing supervision.
System Variables
High staff turnover.
Lack of materials or resources.
Conflicting institutional policies.
The Golden Rule
"Never change an intervention plan based on poor data until you verify that the plan was actually followed."
Fidelity Checklists
How do we objectively measure fidelity?
Identify the critical steps of the intervention.
Create a binary (Yes/No) checklist.
Observer marks completion of each step.
Calculate %: (Steps Done / Total Steps) x 100.
Example: Token Economy Check
1. Identified behavior?
2. Delivered token immediately?
3. Paired with social praise?
4. Used correct ratio?
Fidelity Score: 75%
Treatment Fidelity Teacher Guide Fidelity Facilitator Guide
Lesson 2: Implementation Variables and Fidelity
Executive Summary
The most common error in behavior support is changing a plan because the client's behavior isn't improving, without checking if the staff actually implemented the plan. This lesson teaches students to audit the process before auditing the outcome .
Key Objectives
Identify implementation variables.
Distinguish between intervention and implementation failure.
Calculate treatment fidelity percentages.
Hook Facilitation: The Cookie Mystery
Discussion Prompt: Present the slide. Ask: "If you change the recipe every time the cookies fail, but you keep using salt instead of sugar, will you ever find a good recipe?"
The Clinical Connection:
In the behavior lab, the 'recipe' is the Behavior Intervention Plan (BIP). The 'ingredients' are the reinforcement and prompt procedures. If staff skip steps, the BIP will 'taste' bad (fail), but it might be a perfectly good BIP if followed correctly.
Key Instructional Nodes
1
Intervention Failure vs. Implementation Failure
Intervention Failure: The plan was followed perfectly (90%+ fidelity), but the behavior did not change. (Result: The plan is fundamentally flawed).Implementation Failure: The plan was not followed (e.g., 50% fidelity), and behavior did not change. (Result: We don't know if the plan works or not).
2
The Fidelity Auditor Role
Teach students that a consultant's first job is observation. They shouldn't ask staff "Did you do it?" (Staff will say yes). They must use a Checklist and observe a session to get objective data.
3
Thresholds for Action
Generally, 80-90% fidelity is required to trust data outcomes. Below 80%, the first step is retraining staff, not changing the student's plan.
Facilitating the Simulation
In the student activity, they will read about "The Failing Timeout."
Common Student Mistake:
Students often want to "fix the behavior" by adding more punishment. Steer them toward "fixing the implementation" by training the teacher.
The "Aha!" Moment:
The realization that high student behavior data + low staff fidelity data = a staff training problem, not a student problem.
End of Facilitator Guide — Lesson 2
Fidelity Troubleshoot Activity Case Study: The Failing Plan
Fidelity Troubleshooting Simulation
Investigator:
Case ID:
Background
Student: Leo (Grade 3)
Target Behavior: Out-of-seat behavior during independent work.
The Plan (BIP): Differential Reinforcement of Other Behavior (DRO). Teacher should set a timer for 3 minutes. If Leo stays in his seat, teacher gives him a high-five and a 'sticker token'. If he gets up, teacher resets the timer silently.
The Problem:
After two weeks, Leo’s out-of-seat behavior has actually increased . The teacher wants to cancel the plan and move to a more restrictive setting.
Observation Notes
"10:00 AM: Leo in seat. Teacher busy with another group. 10:05 AM: Leo gets up. Teacher yells 'Leo, sit down!' 10:07 AM: Leo sits. Teacher gives a sticker but doesn't say anything. 10:10 AM: Leo gets up again. Teacher resets timer but also gives him a 'warning look' and 30 seconds of verbal lecturing."
1
Fidelity Checklist Development
Based on "The Plan" above, break down the intervention into 4 critical implementation steps. Then, mark or based on the "Observation Notes."
Critical Step of Intervention Observed?
2
Fidelity Score
Calculation
_____ / 4
Percentage
_______%
3
Failure Type
Based on your audit, is this behavior change (or lack thereof) a result of:
Intervention Failure (Plan is bad)
Implementation Failure (Execution is bad)
4
Consultant Recommendation
What specific steps should you take to adjust this case? (Mention staff training, procedural changes, or data collection needs).
Data Decision Slides The Decision Point
Continue, Modify, or Terminate?
Lesson 03
Choose Your Adventure
You've implemented a plan for 10 days. The data shows:
Success
"Everything is going perfectly."
Stalling
"No change after two weeks."
Crisis
"Behavior is getting worse."
What is your next move?
Universal Decision Rules
1. The "Keep Going" Rule
If data shows progress toward the goal and fidelity is high, DO NOT CHANGE. Resistance to change is common; stay the course.
2. The "3-Data Point" Rule
If three consecutive data points move in the wrong direction, it's time for a "Fidelity Audit" and then a plan modification.
3. The "Ceiling Effect" Rule
If the student has met the goal for 5 days straight, it's time to Fade . Transition to natural supports.
Data-based decisions reduce clinical bias and prevent "program hopping."
The Adjustment Toolbox
Modify
Change the schedule of reinforcement.
Change the prompt level.
Introduce a new motivator.
Fade
Systematically remove the helper.
Increase the time between rewards.
Move to a natural environment.
Terminate
Cease the specialized intervention.
Conduct a final maintenance probe.
Hand off to the natural caregiver.
Simulation: The decision-Maker
In your worksheet, you will act as a Clinical Supervisor for three different cases.
You must interpret the trend, check the fidelity, and decide the fate of the plan. Good luck!
Data Decision Teacher Guide Decision-Making Facilitator Guide
Lesson 3: The Data-Driven Pivot
Executive Summary
Behavioral interventions are dynamic, not static. This lesson moves students away from a "set it and forget it" mentality. They will learn the formal logic for making pivots in treatment, using specific quantitative markers to justify their clinical moves.
Key Objectives
Apply the "3-Data Point Rule" for intervention modification.
Identify master criteria for fading supports.
Justify a termination decision using stability data.
The 3-Data Point Rule
One data point in the wrong direction is a fluke. Two is a concern. Three is a trend.
Instructional Tip: Tell students that if they see 3 consecutive days of worsening behavior, they are ethically obligated to stop and evaluate. They should first check fidelity (Lesson 2), then check the function of the behavior, and finally modify the plan.
Coaching the "Pivot"
The "Continue" Bias
Students often want to change things too quickly because they are anxious for results. Remind them: if the trend is positive (even slowly), stay the course . Rapid changes make it impossible to know what actually worked.
Mastery & Fading
Teach students that "Winning" means the student doesn't need the intervention anymore. If the data is at 100% (or 0% for negative behavior) for 5-10 days, the intervention is now a dependency . Fading must start immediately.
Critical Questioning
"Why is it dangerous to change an intervention based on just one bad day of data?"
Answer: Environmental noise (sickness, bad night's sleep) can cause outliers. Changing the plan based on noise creates "clinical chaos."
"What data point indicates we should terminate the plan entirely?"
Answer: When the student performs at mastery level during maintenance probes (data collection without the intervention in place).
End of Facilitator Guide — Lesson 3
Decision Path Worksheet The Decision Path
Clinical Simulation Workshop
Supervisor:
Quarter:
Your Role: You are the lead clinician reviewing the following three cases. For each case, analyze the provided data trend and fidelity report. Then, circle the most appropriate Action and provide a clinical Justification based on the decision rules discussed in class.
Case 1: Reading Mastery
Fidelity: 95%
10 Days of Progress
Modify Plan
Continue Plan
Fade Plan
Justification (Reference trend & fidelity):
Case 2: Hand-Raising
Fidelity: 90%
2 Weeks - No Change
Modify Plan
Continue Plan
Fade Plan
Justification (Reference trend & fidelity):
Case 3: On-Task Behavior
Fidelity: 100%
Goal Met for 7 Consecutive Days
Modify Plan
Continue Plan
Fade/Terminate
Justification (Reference trend & fidelity):
The Ethics of Stalling
If a client’s behavior isn't changing after 2 weeks of high-fidelity implementation, is it ethical to "wait another week"? Why or why not?
Fading Prompts Slides The Exit Strategy
Fading Prompts and Promoting Independence
Lesson 04
The Teaching Paradox
A clinical supervisor’s success is measured by how unnecessary they eventually become.
Goal 1
Initial Acquisition
"Help them do it right now, no matter how much support it takes."
Goal 2
Prompt Independence
"Remove the help systematically until they do it without you."
Systematic Fading Techniques
Most-to-Least
Starting with full support (physical) and moving to less intrusive (verbal, visual) as behavior improves.
Least-to-Most
Allowing the student a chance to respond independently first, only adding prompts if they fail or delay.
Time Delay
Increasing the time between the instruction and the prompt, encouraging the student to 'beat the clock'.
Thinning Reinforcement
We must move from Continuous to Intermittent reinforcement.
The Natural Schedule
In the real world, we don't get a gold star every time we say 'Please'. We thin the schedule to match natural expectations.
Fading Logic Check
Wait for stability at current level.
Change only ONE variable at a time.
If behavior regresses, move back one step.
Mastery Challenge
Design a fading plan for a peer. How will you go from "Holding their hand" to "Watching from across the room"?
Observe
Thin
Release
Fading Prompts Teacher Guide Independence Facilitator Guide
Lesson 4: Fading and thin reinforcement
Executive Summary
The goal of every intervention is its own termination. This lesson focuses on the clinical skill of "stepping back." Students learn to design fading protocols that prevent prompt dependency —a state where a client can perform a skill but only when specifically prompted by a caregiver.
Key Objectives
Distinguish between prompt types.
Map a "Most-to-Least" fading sequence.
Implement reinforcement thinning schedules.
Hook Facilitation: Peer Skill Transfer
Activity: Have students pair up. One student must teach the other a very specific physical skill they might not know (e.g., a specific card trick, a complex knot, or a pen-spinning trick).
The Instruction:
"Phase 1: You are the shadow. Do the trick with their hands. Phase 2: Give them one verbal hint only. Phase 3: Stand 5 feet away and just watch. If they fail at Phase 3, why did they fail? Did you fade too fast or too slow?"
Clinical Focus Points
1
Identifying Prompt Dependency
Explain that if a student waits for the teacher to point before they open their book, they are prompt dependent . The point is now the "cue" rather than the teacher's instruction. We must fade that point to restore the original cue's power.
2
Thinning vs. Removal
Crucial distinction: We don't just "stop" giving rewards. We thin the schedule. Move from "Every time" (FR1) to "Every 3 times" (FR3) to "Eventually" (Variable Ratio). This creates resistance to extinction.
3
The "One Step Back" Rule
If behavior breaks down during fading, do not go back to the beginning. Go back to the last successful level of support for 2 sessions, then try the fade again with a smaller adjustment.
The Fading Map Workshop
Direct students to use the worksheet to map out a skill. Encourage them to choose complex skills where "Most-to-Least" fading is necessary (e.g., using a communication device, tying shoes, or safety-crossing a street).
End of Facilitator Guide — Lesson 4
Fading Strategy Worksheet The Fading Strategy Map
Designing for Independence
Consultant:
Skill:
Skill Selection & Analysis
Target Skill (e.g., Tying shoes, ordering food):
Natural Stimulus (The "Cue" for the behavior):
1
The Most-to-Least Ladder
Define the progression of prompts from most intrusive (Highest Support) to least intrusive (Lowest Support).
Full Support
1
Prompt Type & Description:
Partial Support
2
Prompt Type & Description:
Minimal Support
3
Prompt Type & Description:
Independent
4
Prompt Type & Description:
No prompts. Performance is triggered solely by the natural stimulus.
2
Reinforcement Thinning Plan
Acquisition Reinforcement (Initial Stage):
e.g., Continuous FR1 - One sticker per attempt
Thinning Goal (Final Stage):
e.g., Intermittent VR5 - One sticker every ~5 attempts
3
The Regression Protocol
If you move from Level 2 to Level 3 on your ladder and the student begins to make consistent errors, what is your immediate clinical response?
Generalization Maintenance Slides The Long Game
Generalization and Maintenance Planning
Lesson 05
The Relapse Mystery
"But he was doing so well in the clinic!"
Scenario A
A child learns to say 'Please' at school, but never says it at home. (Generalization Failure)
Scenario B
A student stops hitting peers in October. By December, the hitting is back. (Maintenance Failure)
If the behavior doesn't last, did it ever really change?
Generalization vs. Maintenance
Generalization
The spread of intervention effects to other settings, people, or behaviors that were not directly treated.
• New settings (Home, Park)
• New people (Dad, Peers)
• New response types
Maintenance
The persistence of the behavior change over time, even after the specialized intervention has been removed.
• Long-term stability
• Resistance to extinction
• Lasting clinical impact
Strategies for Success
1. Train Loosely
Vary the instructions, the materials, and the location. Don't let the student get "locked in" to one specific way of doing things.
2. Natural Contingencies
Ensure the behavior earns rewards in the natural world (e.g., getting a drink after asking) so the environment takes over for the clinician.
3. Program Common Stimuli
Bring elements of the natural setting into the training setting (e.g., teaching playground skills on the actual playground).
The Maintenance Probe
"The Health Check"
A maintenance probe is data collection after the intervention has stopped. It tells you if your work was a 'quick fix' or a 'permanent change'.
"Don't wait for a relapse to find out your maintenance plan failed."
Generalization Maintenance Teacher Guide Long-Term Impact Facilitator Guide
Lesson 5: Generalization and Maintenance
Executive Summary
The final lesson of the sequence shifts the focus from acquisition to endurance . Students will learn that behavior change is not "real" until it occurs in non-training settings and persists after the clinician leaves. This lesson provides the tools for "programming" these outcomes rather than "hoping" for them.
Key Objectives
Identify generalization failures in case studies.
Apply Stokes & Baer’s (1977) generalization strategies.
Design a maintenance probe schedule.
Hook Facilitation: Relapse Autopsy
Case Presentation: Present the scenario of a student who was "cured" of aggression in a psychiatric facility but returned to hitting the moment they got home to their parents.
The Autopsy Question:
"What data points did the facility miss before discharging him? Did they ever test his behavior with his parents? Did they ever test it without the facility's rigid reward system?"
Mastery Content
1
Stokes & Baer Strategy: "Programming"
Emphasize the phrase: "Train and Hope" is not a strategy. If we want behavior to generalize, we must program it. This means intentionally introducing new people and settings midway through the intervention, not at the end.
2
Stimulus vs. Response Generalization
Stimulus: Same behavior, different place (e.g., Saying 'Hi' at school AND the grocery store).Response: Different but related behavior (e.g., Child learns to say 'Hi' and then starts waving and saying 'Hello' on their own).
3
The "Natural Trap"
Teach students to look for behaviors that are "self-reinforcing" in the real world. A child who learns to play with others is "trapped" by the natural fun of play, which maintains the behavior without us needing to give stickers anymore.
The Long-Term Success Plan
The final worksheet asks students to design a 6-month maintenance schedule. Remind them that maintenance probes should be "low stakes"—just observation with no intervention, to see what actually stuck.
End of Facilitator Guide — Lesson 5
Longterm Success Worksheet Long-Term Success Plan
Programming for Generalization and Maintenance
Clinician:
Case:
1
Generalization Planning
Training Setting (Where they learn it):
Natural Setting (Where it matters):
Programming Tactics (Describe 3 specific ways to help the behavior spread):
A
Strategy (e.g., Multiple Exemplars):
B
Strategy (e.g., Programming Common Stimuli):
C
Strategy (e.g., Training Loosely):
2
Maintenance Probe Schedule
Define your follow-up plan for after the intervention has been terminated. How often will you collect "health check" data?
1 Month Post
3 Months Post
6 Months Post
Sequence Synthesis
Reflecting on the entire sequence: If you have high fidelity, a stable baseline, and a successful decision-making process, but you fail to plan for generalization, has the intervention truly succeeded? Justify your stance.
