Percentage of Non-overlapping Data (PND) Calculator

Determine what is used to calculate overlap between conditions in ABA (Applied Behavior Analysis)

What is used to calculate overlap between conditions in ABA?

In the field of Applied Behavior Analysis (ABA), researchers and clinicians frequently ask what is used to calculate overlap between conditions in aba to determine the effectiveness of an intervention. The most widely recognized and utilized metric for this purpose is the Percentage of Non-overlapping Data (PND). PND is a non-parametric method used in single-case experimental designs to quantify the level of change between a baseline phase (Condition A) and an intervention phase (Condition B).

While other methods like NAP (Non-overlap of All Pairs) or PEM (Percentage Exceeding the Median) exist, PND remains the clinical standard. It specifically looks at how many data points in the treatment phase exceed the most extreme data point in the baseline phase. This provides a clear, numerical effect size that helps clinicians decide if a behavior plan is working or needs adjustment.

Who Should Use This Metric?

• Board Certified Behavior Analysts (BCBAs): To validate the efficacy of behavior intervention plans.
• Special Educators: To track progress in IEP goals using single-subject data.
• Researchers: To report effect sizes in peer-reviewed ABA literature.

Formula and Mathematical Explanation

To understand what is used to calculate overlap between conditions in aba, one must master the PND formula. The calculation is straightforward but requires identifying the “extreme” baseline point based on the clinical goal.

The PND Formula:

PND = (Number of Intervention Points > Baseline Maximum / Total Number of Intervention Points) × 100

Note: If the goal is to decrease a behavior, the calculation uses points lower than the baseline minimum.

Variable	Meaning	Unit	Typical Range
Phase A Max/Min	The highest (for increase) or lowest (for decrease) baseline point.	Count/Time	Varies by behavior
N (Overlap)	Count of treatment points that don’t overlap with baseline.	Integer	0 to Total Points
Total Points	Total number of observations in the intervention phase.	Integer	Minimum 3-5
PND %	The final calculated overlap percentage.	Percentage	0% to 100%

Scruggs and Mastropieri Criteria

When you determine what is used to calculate overlap between conditions in aba, you need a way to interpret the results. The Scruggs and Mastropieri (1998) scale is the gold standard:

• 90% or higher: Highly effective intervention.
• 70% to 90%: Moderately effective.
• 50% to 70%: Questionably effective.
• Below 50%: Ineffective.

Practical Examples (Real-World Use Cases)

Example 1: Increasing Functional Communication

A child uses “mands” (requests) during baseline at frequencies of 2, 4, 3, 5, and 4. The maximum baseline point is 5. After starting a PECS intervention, the data points are 6, 8, 5, 9, 10, 12.

Overlapping points: 1 (the ‘5’).

Non-overlapping points: 5 (6, 8, 9, 10, 12).

Calculation: (5 / 6) × 100 = 83.3%. This is a moderately effective intervention.

Example 2: Decreasing Self-Injurious Behavior (SIB)

Baseline instances of SIB: 15, 12, 18, 20. The minimum baseline point is 12. Treatment starts: 10, 8, 5, 11, 13.

Non-overlapping points (those below 12): 4 (10, 8, 5, 11).

Calculation: (4 / 5) × 100 = 80%. This shows moderate success in reducing the behavior.

How to Use This PND Calculator

1. Enter Baseline Data: Type in your Phase A data points separated by commas.
2. Enter Intervention Data: Type in your Phase B data points.
3. Select Goal: Choose whether you want the behavior to increase or decrease.
4. Analyze Graph: Look at the visual chart to see how the intervention phase compares to the baseline extreme line.
5. Review Results: Check the PND percentage and the interpretation based on clinical standards.

Key Factors That Affect Overlap Calculation Results

1. Baseline Volatility: A single outlier in the baseline (e.g., a huge spike) can drastically lower the PND result, making an effective treatment look ineffective.
2. Data Path Trend: PND does not account for existing trends. If behavior was already improving during baseline, PND might overestimate intervention success.
3. Phase Length: Having too few points in either phase (less than 3) makes the calculation statistically weak.
4. Ceiling/Floor Effects: If baseline data already reaches near-zero or maximum levels, calculating overlap becomes difficult.
5. Inter-observer Agreement (IOA): The reliability of the raw data itself determines if the PND result is trustworthy.
6. Measurement Sensitivity: Whether you use whole-interval, partial-interval, or frequency recording affects the range of numbers being calculated.

Frequently Asked Questions (FAQ)

Q: What is the most common tool used to calculate overlap between conditions in ABA?
A: The Percentage of Non-overlapping Data (PND) is the most common visual analysis tool used for this purpose.

Q: Can PND be used for decreasing behaviors?
A: Yes. For decreasing behaviors, we calculate the percentage of treatment points that are lower than the lowest baseline point.

Q: What are the limitations of PND?
A: Its biggest limitation is sensitivity to baseline outliers. One high baseline point can result in a 0% PND even if the intervention clearly changed the behavior.

Q: Is PND the same as effect size?
A: It is a type of non-parametric effect size specifically designed for single-case designs, unlike Cohen’s d which is used for group designs.

Q: How many data points are needed for a valid PND?
A: While there is no hard rule, ABA standards generally suggest at least 3-5 data points per phase for stable analysis.

Q: Why not just use visual analysis?
A: Visual analysis is subjective. PND provides a concrete number that researchers can use to compare different studies.

Q: Does PND work if the baseline has a trend?
A: Not well. If the baseline has an improving trend, the Percentage of Data Exceeding the Median (PEM) or Tau-U might be more appropriate.

Q: What is a “good” PND score?
A: Generally, anything above 90% is considered highly effective in a clinical ABA setting.

Related Tools and Internal Resources

• IOA Calculator – Calculate Inter-observer Agreement for ABA sessions.
• ABA Graphing Templates – Excel and Google Sheets templates for Phase A/B charting.
• Preference Assessment Tool – Identify reinforcers before starting intervention.
• Trial-by-Trial Data Tracker – Record discrete trial training data easily.
• VB-MAPP Progress Tracker – Align your overlap calculations with milestone goals.
• ABC Data Analyzer – Analyze antecedent-behavior-consequence patterns.