A. Measurement Flashcards
A. Measurement
Measuring Behavior
Measuring observable behavior is a hallmark of Applied Behavior Analysis and includes the precise measurement of responses to changes in the environment.
Preparing for Data Collection
To prepare for data collection first you must know and understand the behavior you are recording. Typically, your supervisor will have defined the behavior you are recording and will have created a data sheet outlining the parameters in which data should be recorded. When preparing direct measurement recording (in addition to a writing utensil and data sheet) you may need:
* A timer to record duration
* A clicker to record frequency
* Appropriate stimuli
A. Measurement
Preparing for Data Collection
Data collection is an important part of an RBTs job. Supervisors and families rely on RBTs to collect strong data to help make programming decisions. Without strong data collection no one will be able to determine if a client is making progress. Data collection allows us to determine when and if a client is making progress.
Task Analysis of Data Collection
1. The type of behavior you are collecting data on: such as duration, rate, interval
2. Determine the best data collection method. Your supervisor will help. This is often outlined in the functional behavior assessment or the behavior intervention plan
3. Gather the materials you need to collect data: paper, pen, clipboard, timer, data collection sheet, and sometimes an electronic method
4. Fill out any identifying information needed. This could include name, birthdate, diagnosis, and the behaviors or skills being observed. If there is no standard fill-in form, write it at the top of your data sheet
5. Try to eliminate distractions
6. Begin collecting data
A. Measurement
Direct vs. Indirect Measurement
As a behavior technician you will be recording direct observations of behavior majority of the time as direct observation data is the most accurate.
Direct Measurement is recording behavior as it occurs in either natural or contrived settings
* Direct measurement can be continuous or discontinuous
* Recording would be frequency, duration, latency, intensity
* Discontinuous measurement is a sample of possible occurrences of behavior
Indirect measurement is when a target behavior is not directly observed, conclusions are drawn from other behaviors or environment events. Types of indirect measurement include:
* Permanent product recording, self-monitoring, interviews, surveys and rating scales.
A. Measurement
Continuous Measurement Procedures
Continuous measurement means measuring each and every instance of behavior within the observation period. In other words, you miss nothing. This type of measurement is very time consuming and more difficult than discontinuous measurement procedures.
- Frequency: This is a simple count of instances of behavior, represented by a tally. How many times did John hit another student and present the count as a number. John hit another student 5 times
- Rate: Rate is a frequency count with a time element. If you are tallying how many times John hits another student you would also report the time. John hits at the rate of 5 times per hour. Rate is an important measurement when looking at behaviors which are frequent and short. For example: hitting, raising hand, flapping hands, disrupting another student, yelling
- Duration: This is how long a behavior occurs. To take duration data, you start a stopwatch when the behavior begins and end the stopwatch when the behavior stops. Duration is often reported as an average over time. Duration recording is appropriate for long lasting behaviors. For example: tantrums, social play, how long it takes a child to get dressed
A. Measurement
A-02 Implementing continuous measurement procedures:
Frequency: Rate = Count/Time Unit
Frequency recording entails continuously recording each separate instance of behavior
To record frequency you count the behavior when it occurs
Behaviors that you would typically use frequency to record are: hitting, kicking, throwing objects, cursing, pressing buttons, spinning, eating a nonfood item, placing items in a container, etc.
* Recording can be done with a tally mark, numbers or check marks or any symbol that indicates an instance has occurred
A. Measurement
Continuous Measurement Procedures
Inter Rate Response: This is the time between responses. To take IRT data you start the stopwatch when the behavior ends and stop the stopwatch when the behavior begins again. IRT is typically reported as an average. IRT is appropriate when the time between behavior is important. For example: time lapse between prosocial behaviors.
* Latency: Latency is the time from prompt to the start of the behavior. To take latency data start the stopwatch when the behavior starts. You might want to take latency data when there is a delay between the prompt and when the behavior occurs. For example, the time from a prompt to get dressed to a person getting dressed, the time from the instruction to begin a math problem and the response
A. Measurement
A-02 Implementing continuous measurement procedures:
Duration - Recorded per event or “response cycle”
* Duration recording records the entire amount of time the target response occurs
* We record duration when behaviors vary in length of time
* Responses recorded should be continuous, for example,
* Amount of time doing homework
* Thumb sucking
* Sitting in chair
* Singing
* Time out of seat during a lesson
* Duration is easily recorded with a stopwatch or the timer on a cell phone
Behavior must be precisely defined so that you know when you should recording and the time recording should end
A. Measurement
Discontinuous Measurement Procedures
Discontinuous measurement procedures are samples of the target behavior, but they do not measure every instance of behavior within the observation period. Discontinuous measurement procedures are used when it is too time consuming to take continuous measurement data
* Partial Interval: Recording the presence or absence of a behavior during a brief interval of time. Intervals are marked as “+” if the target behavior occurred at any time during the interval. Intervals are marked as “-“ if the target behavior did not occur during the entire interval. Partial interval data collection overestimates the occurrence of behavior. Examples of behaviors that are appropriate for partial interval recording are vocal stereotypy, hand flapping, biting nails.
A. Measurement
A-02 Implementing continuous measurement procedures
- Latency - Time between stimulus and response
- Latency recording records the amount of time that lapses from the onset of a stimulus to the target response
- Latency is recorded when the behavior of interest needs to begin in a specific amount of time
- Examples of behavior where latency is appropriate to record:
- Amount of time between instruction and compliance
_ Amount of time between the bell ringing and child enters class - Latency is easily recorded with a stopwatch or the timer on a cell phone
- Behaviors must be precisely defined so that you know when to begin recording and when the time recording should end
A. Measurement
Discontinuous Measurement Procedures
- Whole Interval: Recording the presence or absence of a behavior during a brief interval of time.
Intervals are marked as “+” if the target behavior occurred during the whole interval. Intervals are marked as “-“ if the target behavior stopped at any time during the interval. Whole interval recording underestimates the occurrence of behavior. Examples of behaviors that are appropriate for whole interval interval recording are cooperative play, social engagement, on task behavior. - Momentary Time Sampling: Recording the presence or absence of a behavior at the very end of an interval. Intervals are marked as “+” if the target behavior occurred at the end of the interval. Intervals are marked as “-“ when the target behavior does not occur at the end of the interval. This procedure is possible to do for many clients at the same time. This is the easiest type of discontinuous measurement procedure to use. However, it provides the least amount of information. Examples of behaviors that are appropriate for this procedure recording are social engagement of many students and task engagement for a group of students
A. Measurement
A-03 Implementing discontinuous recording procedures
Interval Recording: Provides an estimate of behaviors
* Sessions are divided into time periods that are equal
* Each interval is scored as an (+) or non-occurrence (0) according to criteria defined by the supervisor.
* There are three types of interval recording:
- Partial Interval
_ Whole Interval
- Momentary Time Sampling
* Whole interval recording is used to determine whether or not a behavior occurred for an ongoing amount of time. Use if the goal is to increase behavior. Whole interval recording underestimates behavior
- Divide intervals into equal intervals (for example, 10 seconds) in whole interval recording only put a + if the behavior occurred entire length of the interval
* Examples of whole interval recording situations:
- On task behavior
- Sitting in seat
A. Measurement
Permanent Product Procedures
Permanent product recording is not recording behavior but recording the products behavior produces. For example, you could record how many questions a student answered on a worksheet by simply looking at the worksheet and counting the problems completed. The advantage is you can record information from permanent product anytime. The disadvantage is you are not actually recording behavior. For example, a parent could complete a math worksheet for a child.
Examples of Permanent Product Recordings:
1. How many items were placed on a shelf
2. How much homework was completed
3. How many bracelets were constructed
4. How many dished cleaned
5. How many scratches a person has
A. Measurement
A-03 Implementing discontinuous measurement procedures:
- Partial interval recording is used to determine whether or not a behavior occurred within a certain time period. Use if the goal is to decrease a behavior. Partial interval recording overestimates behavior.
- Procedures for using partial interval recording
- Divide intervals into small equal intervals (for example, 10 seconds) in partial interval recording put a + if the behavior occurred at anytime during the interval and a0 if the behavior did not occur
- Examples of partial interval recording situations
- Aggressive behavior
- Scripting
- Calling out
A. Measurement
Enter Data and Update Graphs
Recording data is an important part of an RBTs job. You should receive directions from your supervisor on how to accurately collect and graph data. Graphing can be done with a pencil and paper or a computer program.
How to Summarize Different Types of Data
1. Frequency is summarized rate over sessions
2. Duration is summarized as total duration over sessions
3. IRT is summarized as an average
4. Latency is summarized as latency to response
5. Interval data is summarized as percent intervals with occurrence
A. Measurement
A-03 Implementing discontinuous measurement procedures
- Momentary Time Sampling is used to determine whether or not a behavior occurs during a time period.
- Procedures for using momentary time sampling
- Divide intervals into small equal intervals (for example, 10 seconds) at the end of the set interval observe the individual and record whether or not the behavior is occurring
- Examples of momentary time sampling recording situations
- When you cannot attend to client the entire time
- Behavior has no clear beginning or end
- Behavior with high rates
