RBT Measurement Guide: Master Data Collection

Measurement forms the foundation of applied behavior analysis, providing the objective data necessary to make informed clinical decisions. As an RBT, proficiency in various measurement procedures is essential for accurately documenting behavior and evaluating intervention effectiveness.

A-1: Prepare for Data Collection

Types of Measurement Systems

Continuous Measurement

• Definition: Measurement systems that record each instance of behavior as it occurs
• Common Types:
  • Event recording
  • Frequency counting
  • Rate recording
  • Duration recording
  • Latency recording
  • IRT (Inter-response time) recording
• When to Use: Best for behaviors with clear beginning/end that occur at moderate rates
• Example: Recording each instance of hand-raising during a 30-minute class period

Discontinuous Measurement

• Definition: Measurement systems that sample behavior during specified time intervals
• Common Types:
  • Momentary time sampling
  • Partial interval recording
  • Whole interval recording
• When to Use: Best for high-frequency behaviors or behaviors that occur for extended durations
• Example: Recording if self-stimulatory behavior is occurring at the end of each 1-minute interval

Permanent Product Recording

• Definition: Measurement of behavioral outcomes or products that remain after the behavior
• Common Applications:
  • Completed worksheets
  • Constructed items
  • Written responses
• When to Use: When direct observation is not feasible or when evaluating the quality of behavioral outputs
• Example: Counting completed math problems on a worksheet

Selecting Appropriate Measurement Systems

The choice of measurement system should be based on:

1. Behavior Characteristics: Frequency, duration, intensity
2. Environment Constraints: Available observation time, setting limitations
3. Measurement Purpose: Progress monitoring, baseline establishment, etc.
4. Resources Available: Number of observers, technology access

Decision-Making Flow Chart:

1. Can the behavior be easily counted as discrete events? → Consider event recording
2. Does the behavior occur at very high rates? → Consider time sampling
3. Is the behavior’s duration more important than frequency? → Consider duration recording
4. Is observer continuous availability limited? → Consider permanent product or discontinuous methods

Preparing Materials for Data Collection

Essential components for effective data collection include:

• Operational Definitions: Clear, objective descriptions of target behaviors
• Data Sheets: Properly formatted for selected measurement system
• Timing Devices: Watches, timers, or electronic applications
• Organizational Systems: Client files, data collection apps, etc.

Common Preparation Errors to Avoid:

• Imprecise operational definitions leading to inconsistent measurement
• Inappropriate matching of measurement systems to behaviors
• Inadequate space on data sheets for recording
• Failure to have backup systems for technology failures

A-2: Implement Continuous Measurement Procedures

Event Recording

Implementation Steps:

1. Clearly define the target behavior with observable parameters
2. Determine the observation period length
3. Tally each instance of behavior as it occurs
4. Calculate rate (count ÷ time) if applicable

Example Data Sheet Format:

Client: _______ Date: _______ Observer: _______
Behavior: Hand-raising (defined as extending arm upward with hand open)
Session Time: 9:00-9:30 (30 minutes)
Tally: IIII IIII II (12 instances)
Rate: 0.4 responses per minute (12 ÷ 30)

Common Challenges and Solutions:

• Challenge: Missing instances during high-frequency occurrences
• Solution: Consider using mechanical counters or switching to time sampling
• Challenge: Determining if closely spaced occurrences are separate instances
• Solution: Establish clear criteria for what constitutes a new instance

Duration Recording

Implementation Steps:

1. Define behavior start and end criteria precisely
2. Start timing when behavior begins
3. Stop timing when behavior ends
4. Record total duration
5. For multiple occurrences, sum all durations or calculate percentage of session time

Example Data Sheet Format:

Client: _______ Date: _______ Observer: _______
Behavior: On-task behavior (defined as eyes on work materials, appropriate manipulation of materials)
Session: 10:00-10:30 (30 minutes = 1800 seconds)
Duration Intervals:
Start: 10:00 - Stop: 10:08 (480 seconds)
Start: 10:12 - Stop: 10:22 (600 seconds)
Start: 10:25 - Stop: 10:30 (300 seconds)
Total Duration: 1380 seconds
Percentage of Session: 76.7% (1380 ÷ 1800)

Common Challenges and Solutions:

• Challenge: Determining exact moment behavior starts/stops
• Solution: Create precise definitions with examples of transition moments
• Challenge: Managing timing while monitoring multiple behaviors
• Solution: Use technology aids or consider alternative measurement systems

Latency Recording

Implementation Steps:

1. Define the stimulus presentation clearly
2. Define the target response precisely
3. Start timing at stimulus presentation
4. Stop timing at initiation of response
5. Record the elapsed time (latency)

Example Data Sheet Format:

Client: _______ Date: _______ Observer: _______
Stimulus: Verbal instruction "Please sit down"
Response: Sitting (defined as contact of buttocks with chair seat)
Trial 1: 5 seconds
Trial 2: 3 seconds
Trial 3: 7 seconds
Trial 4: 2 seconds
Trial 5: 3 seconds
Average Latency: 4 seconds

Common Challenges and Solutions:

• Challenge: Inconsistent stimulus presentation
• Solution: Script or standardize stimulus delivery
• Challenge: Partial or incomplete responses
• Solution: Clearly define criteria for response completion

A-3: Implement Discontinuous Measurement Procedures

Momentary Time Sampling

Implementation Steps:

1. Divide observation period into equal intervals
2. At the end of each interval, instantly observe if behavior is occurring
3. Record occurrence (+) or nonoccurrence (-)
4. Calculate percentage of intervals with behavior

Example Data Sheet Format:

Client: _______ Date: _______ Observer: _______
Behavior: Reading (defined as eyes tracking text on page)
Interval Length: 1 minute
Interval   1   2   3   4   5   6   7   8   9   10
Recording  +   +   -   -   +   +   -   +   +   -
Percentage of Intervals with Behavior: 60% (6/10)

Advantages:

• Less observer effort than continuous recording
• Minimizes observer bias toward noticing behavior at other times
• Generally underestimates behavior occurrence (conservative)

Disadvantages:

• May miss behavior that occurs between observation points
• Less precise than continuous measurement
• Can significantly under/overestimate if interval size is inappropriate

Partial Interval Recording

Implementation Steps:

1. Divide observation period into equal intervals (typically 10-30 seconds)
2. Record if behavior occurs at any point during each interval
3. Behavior is scored as present (+) if it occurs at any time in interval
4. Calculate percentage of intervals with behavior

Example Data Sheet Format:

Client: _______ Date: _______ Observer: _______
Behavior: Out-of-seat (defined as no buttocks contact with chair)
Interval Length: 15 seconds
Interval   1   2   3   4   5   6   7   8   9   10
Recording  +   +   -   -   +   -   -   +   -   -
Percentage of Intervals with Behavior: 40% (4/10)

Advantages:

• Captures behaviors that occur at any point during intervals
• Useful for measuring behaviors of shorter duration
• Generally overestimates actual occurrence (conservative for problem behaviors)

Disadvantages:

• Overestimates actual duration of behavior
• Cannot distinguish between one long occurrence and multiple brief occurrences
• More effort than momentary time sampling

Whole Interval Recording

Implementation Steps:

1. Divide observation period into equal intervals
2. Behavior must occur throughout entire interval to be scored
3. Record as occurrence (+) only if behavior present for full interval
4. Calculate percentage of intervals with continuous behavior

Example Data Sheet Format:

Client: _______ Date: _______ Observer: _______
Behavior: Engagement (defined as manipulation of task materials)
Interval Length: 10 seconds
Interval   1   2   3   4   5   6   7   8   9   10
Recording  +   -   -   +   +   -   +   -   -   -
Percentage of Intervals with Behavior: 40% (4/10)

Advantages:

• Useful for measuring sustained behavioral states
• Generally underestimates behavior (conservative for appropriate behaviors)
• Good for behaviors that need to occur continuously to be functional

Disadvantages:

• Underestimates actual occurrence
• Requires constant vigilance throughout each interval
• May fail to capture brief but important instances of behavior

A-4: Implement Permanent Product Recording Procedures

Types of Permanent Products

Academic Products:

• Completed worksheets
• Written assignments
• Test responses
• Art projects

Behavioral Products:

• Cleaned areas
• Constructed items
• Organized materials
• Digital creations

Data Collection Considerations:

1. Product must have measurable dimensions
2. Product must remain available for assessment
3. Connection between behavior and product must be clear

Measurement Dimensions

Accuracy Measurement:

• Count correct vs. incorrect responses
• Calculate percentage correct (correct ÷ total × 100)
• Compare against mastery criteria

Rate-Based Measurement:

• Count number of products completed
• Calculate rate (count ÷ time)
• Evaluate against fluency criteria

Quality Assessment:

• Use rating scales with operational definitions
• Employ rubrics with specific criteria
• Ensure inter-rater reliability when subjective

Example Data Sheet Format:

Client: _______ Date: _______ Assessor: _______
Product: Math worksheet
Items Attempted: 20
Items Correct: 16
Accuracy: 80% (16 ÷ 20 × 100)
Completion Time: 15 minutes
Rate: 1.33 problems per minute (20 ÷ 15)
Quality Ratings (1-5 scale):
  Neatness: 4
  Show of Work: 3
  Following Format: 5

Implementation Challenges and Solutions:

• Challenge: Determining if product was completed independently
• Solution: Implement supervision/monitoring procedures
• Challenge: Distinguishing between product quality and behavior quality
• Solution: Define specific product attributes tied to behavioral targets

A-5: Enter Data and Update Graphs

Accurate Data Entry

Essential Practices:

1. Enter data promptly after collection
2. Use consistent formats for dates, times, and metrics
3. Include session information (observer, setting, conditions)
4. Double-check entries against raw data
5. Maintain data security and confidentiality

Common Data Entry Errors:

• Transposing numbers
• Calculating percentages incorrectly
• Using inconsistent measurement units
• Missing data points without notation
• Confusing dates or session numbers

Creating and Updating Graphs

Graph Types and Applications:

• Line Graphs: Most common, showing behavior change over time
• Bar Graphs: Comparing discrete categories or conditions
• Cumulative Graphs: Displaying total performance over time
• Scatterplots: Examining relationships between variables

Essential Graph Elements:

1. Title: Clear description of behavior and client (use initials for confidentiality)
2. Axes Labels: Specific behavioral measures and time units
3. Phase Change Lines: Vertical lines indicating intervention changes
4. Phase Labels: Brief descriptions of conditions
5. Data Points: Consistently formatted symbols
6. Scale: Appropriate to capture behavior range

Example Line Graph Construction:

1. Determine appropriate scale based on data range
2. Plot each data point accurately
3. Connect data points within same phase/condition
4. Add phase change lines at intervention transitions
5. Label phases clearly
6. Include legend if multiple behaviors or conditions

Electronic vs. Hand-Drawn Graphing:

• Electronic advantages: easy updating, professional appearance, automatic calculations
• Hand-drawn advantages: immediate availability, no technology requirements
• Both require accuracy and attention to detail

Graphing Errors to Avoid:

• Inconsistent scales that distort visual analysis
• Missing or unclear labels
• Connecting data points across phase changes
• Inappropriate graph types for the data
• Cluttered presentations that obscure patterns

A-6: Describe Behavior and Environment

Objective vs. Subjective Description

Objective Description (Required):

• Based on observable, measurable characteristics
• Free from interpretation or inference
• Consistent across observers
• Specific and precise

Subjective Description (Avoid):

• Based on opinions or judgments
• Contains interpretive language
• May vary between observers
• Often uses vague terminology

Examples Contrasting Objective vs. Subjective:

Operational Definitions

Components of an Effective Operational Definition:

1. Objective Description: What the behavior looks like
2. Temporal Parameters: Beginning and end of the behavior
3. Intensity Threshold: Minimum level to count as occurrence
4. Exclusionary Criteria: Similar behaviors that should not be counted
5. Examples and Non-examples: Clear illustrations of the definition in practice

Examples of Strong Operational Definitions:

Tantrum: Vocalizations above conversational level accompanied by crying, falling to floor, and/or physical contact with objects with force sufficient to make an audible sound, lasting for at least 3 seconds. A new instance is recorded after 5 seconds of calm behavior. Does not include whining at conversational level or brief expressions of frustration without physical components.

On-Task Behavior: Eyes oriented toward assigned work materials or teacher during instruction, manipulating materials as designed for their intended instructional purpose, and responding verbally to teacher queries within 5 seconds. Excludes manipulation of materials without academic engagement (e.g., doodling, stacking materials) and looking at instructional materials without active engagement for more than 10 seconds.

Descriptor Types for Environmental Variables

Example Environmental Description: Observation occurred in the classroom during small group math instruction. The client was seated at a table with three peers. The room was approximately 70°F with overhead fluorescent lighting. Noise level was moderate with other groups working nearby. The task involved double-digit multiplication problems of moderate difficulty. The instructor was providing intermittent verbal praise for work completion and proximity (within 3 feet) approximately every 2 minutes.