Temporal Dimensions of Behavior
Applied behavior analysis measures behavior across multiple dimensions. While frequency and rate are perhaps the most commonly used measures, many clinically important behaviors are best understood through their temporal properties: how long they last, how quickly they begin, or how much time passes between occurrences. Latency, duration, and inter-response time (IRT) are three temporal measurement procedures that capture these different aspects of behavior.
Understanding when to use each measure and what information it provides is essential for both clinical practice and the BCBA exam. Choosing the wrong measurement procedure can lead to inaccurate data, poor clinical decisions, and missed intervention effects.
Latency
Definition
Latency is the elapsed time from the onset of a stimulus (such as an instruction, a signal, or an environmental event) to the onset of the response. It measures how quickly an individual begins responding after a relevant stimulus occurs.
When to Use Latency
Latency is the appropriate measure when the speed of response initiation is the primary concern. Common clinical applications include:
- Compliance: Measuring how quickly a child begins following an instruction after it is given. A child might eventually comply with every request but take 10 minutes to start, which is a latency issue, not a frequency issue.
- Task initiation: Measuring how long it takes a student to begin working after the teacher gives directions.
- Social responses: Measuring how quickly an individual responds to a social greeting or conversational turn.
- Safety behaviors: Measuring how quickly someone evacuates after a fire alarm sounds.
How to Measure
To measure latency, start a timer when the stimulus occurs and stop it when the response begins. Record the elapsed time for each opportunity. Data can be reported as individual latency values, mean latency across opportunities, or median latency.
Accurate latency measurement requires precise identification of both the stimulus onset and the response onset. Operational definitions should clearly specify what counts as the stimulus (for example, the end of a verbal instruction) and what counts as the start of the response (for example, the individual's first movement toward the task).
Example Calculation
A teacher gives five instructions during a session. The student begins complying after 3 seconds, 8 seconds, 2 seconds, 15 seconds, and 4 seconds. The mean latency is (3 + 8 + 2 + 15 + 4) / 5 = 6.4 seconds. If the intervention goal is a latency of 5 seconds or less, the mean latency is slightly above criterion, and the individual high latency of 15 seconds suggests inconsistency that may warrant further analysis.
Duration
Definition
Duration is the elapsed time from the onset of a response to its offset. It measures how long a behavior lasts once it begins. Duration can be reported as total duration (cumulative time spent engaged in the behavior during an observation period) or per-occurrence duration (how long each individual instance of the behavior lasts).
When to Use Duration
Duration is appropriate when the length of time a behavior persists is the primary concern. Clinical applications include:
- Tantrums: A child may have only one tantrum per day (low frequency), but if it lasts 90 minutes, duration is the more meaningful measure.
- On-task behavior: Measuring how long a student sustains attention to academic materials.
- Exercise: Measuring how long a client engages in physical activity.
- Sleep: Measuring total sleep duration for clients with sleep difficulties.
- Stereotypy: Measuring the total time spent engaged in repetitive behaviors.
How to Measure
Start a timer when the behavior begins and stop it when the behavior ends. For total duration, use a cumulative timer that runs during all instances of the behavior throughout the observation period. For per-occurrence duration, record the duration of each individual instance separately.
Total duration: The sum of all durations during the observation. For example, if a student was on-task for three intervals of 5 minutes, 8 minutes, and 3 minutes during a 30-minute observation, the total duration of on-task behavior is 16 minutes.
Per-occurrence duration: The duration of each individual episode. Using the same example, the per-occurrence durations are 5, 8, and 3 minutes, with a mean per-occurrence duration of 5.3 minutes.
Percentage of total duration: Duration is often expressed as a percentage of the total observation time. In the example above, 16 minutes / 30 minutes = 53.3% of the observation period.
Inter-Response Time (IRT)
Definition
Inter-response time (IRT) is the elapsed time between the end of one response and the beginning of the next response. It measures the temporal spacing between consecutive instances of the same behavior.
When to Use IRT
IRT is appropriate when the spacing between responses is clinically relevant. Applications include:
- Pacing: A student who rushes through math problems (short IRT) may benefit from an intervention that increases the time between responses to improve accuracy.
- Conversational turn-taking: Measuring the time between one person finishing a statement and the other person responding helps evaluate social communication skills.
- Repetitive behaviors: Measuring the time between episodes of stereotypy or self-injury can reveal patterns related to environmental events.
- Schedule thinning: Monitoring IRT during differential reinforcement of low rates (DRL) schedules, where reinforcement is delivered only when the IRT exceeds a specified minimum.
How to Calculate
IRT is calculated by subtracting the time of the previous response's offset from the time of the next response's onset. For example, if a client finishes one request at 2:00 and begins the next request at 2:45, the IRT is 45 seconds.
IRT data are often reported as mean IRT, but the distribution of IRTs can also be informative. A wide range of IRTs may suggest that the behavior is influenced by different variables at different times.
Relationship to Other Measurement Procedures
Latency, duration, and IRT are related to other commonly used measures. Rate (frequency divided by time) captures how often a behavior occurs but not how long each instance lasts. Frequency counts instances but does not capture temporal properties. These temporal measures complement frequency and rate by providing a more complete picture of the behavior.
In clinical practice, behavior analysts often need to collect multiple types of data simultaneously to capture all relevant dimensions. For example, measuring both the frequency and duration of tantrums provides a more complete picture than either measure alone.
BCBA Exam Scenarios
On the exam, you may encounter scenarios asking you to select the most appropriate measure for a given behavior. Remember these decision rules: Choose latency when the concern is how quickly behavior starts after a stimulus. Choose duration when the concern is how long the behavior persists. Choose IRT when the concern is the time between consecutive responses. Choose frequency or rate when the concern is how often the behavior occurs regardless of its temporal properties.
Practice identifying the most relevant dimension for different behaviors, and be prepared to explain why one measure is more appropriate than another in a given clinical scenario.