Skill Repertoire Building Data Collection

Introduction to Data Collection

This lesson on Data Collection will introduce you to the methodologies and procedures used to collect data during the Skill Repertoire Building Components of a Therapy Session. Data Collection is a very important element of ABA Therapy. The data collected, is used by the BCBA to make important decisions about a child’s program. Because of this, it is critical that all therapists are consistent in their data collection and are producing reliable data. The specific forms to be used are reviewed in subsequent lessons.

Purpose and Guidelines for Data Collection in Skill Repertoire Building:

Data Collection is the observation of a Target Behavior or skill, the measurement of that behavior, and a permanent record of the measured behavior. For example, when teaching a child to name objects, data collection requires that the therapist must observe the Target Behavior (the child naming objects), measure the behavior (number of times the child names objects), and create a permanent record (record that number on a data sheet).

Data collection is a critical component of any behavior analytic intervention used to teach a child new skills. This is because data collection allows the program supervisor (BCBA) to make data-based decisions regarding a child’s programming. Through data collection the BCBA is able to identify skill deficits, monitor progress, determine mastery and ensure maintenance and generalization of a skill.

Identifying skill deficits requires that data is collected before teaching a new skill in order to determine which skills need to be taught and which skills a child already knows and can perform independently.

Data is also collected while teaching a new skill in order to monitor the child’s progress, to determine if the instructional strategy is successful, to identify when a child is ready to move to a new step of discrimination training, to determine when a new step in a behavioral chain is needed or to determine when mastery has been achieved.

Data collection is also used after a skill is learned in order to ensure that the skill maintains over time and generalizes to novel circumstances.

Why Reliable Data Collection is Important:

With the data provided from data collection the program supervisor (BCBA) is able to make good decisions about what skills to teach, how to teach them, when and where they should be taught and to determine when a skill has been mastered. Because the program supervisor uses the collected data to guide decisions regarding a child’s program, it is extremely important to have reliable data. Therefore the data must be accurate and consistently collected.

There are two guidelines for ensuring reliable data collection. First, data should be recorded immediately following the occurrence of behavior. When there is a delay between the occurrence of the behavior and when the therapist records the data it is more likely that the therapist may make errors or even forget to record data all together. For example, when teaching a child to answer social questions, such as “what is your name” or “how old are you”, the therapist must record the child’s response immediately or risk forgetting what the child said or whether the answer was correct or incorrect.

The second guideline for data collection is what data should be collected. In other words all relevant information regarding the observation and measurements of the behavior should be recorded to provide a complete account of the target skill. In returning to the example of a child answering social questions,  in addition to recording data on the child’s response, the therapist may also need to take data on the SdS presented, prompts provided, the current step of discrimination training and the date the skill was practiced.

By following these guidelines, the data collection is more likely to be accurate and reliable. And the program supervisor (BCBA) will be able to make decisions that will be most beneficial for the child.

Inter-observer Agreement & How it is Used to Determine the Reliability of Data Collection:

One way to assess the reliability of data is the Interobserver Agreement (IOA). Inter observer agreement measures data reliability by calculating the degree to which two people agree on a behavior that they have both observed, measured and recorded. Practically speaking IOA is calculated by having two people independently observe, measure and record the same behavior. The data they collect is then compared, to determine the level of similarity or “agreement”.

When the collection results in identical or very similar sets of data the data is considered to be reliable. However, if the level of agreement between the two observers is low and the respected data collected results in very different sets of data than the data is considered to be unreliable, and the program supervisor is unable to use the data to make informed decisions.

Example:

If two therapists observe a child answering social questions and one therapist records 10 correct responses and the other records 9 correct responses this data is similar and considered reliable. Alternately, if one therapist records 10 correct response and the other records 3 correct response than the data is considered unreliable.

In the first case the program supervisor can be confident that the child knows the answers to the social questions because both therapists agree that the child made several correct responses. In the second case the program supervisor is unable to determine whether the child has learned the skill or not because the data is not reliable. As can be seen, IOA is an effective tool for determining whether behavior is being measured and recorded accurately and consistently.

Two Types of Recording Methods:

A effective skill building program must include a detailed description of the recording method used to collect data. The recording method specifies the observation period and when to record data.

The Observation Period: Is a specific time frame when the behavior is likely to occur. And the therapist should be prepared to observe and measure it.

We’ll take a look at two methods commonly used in skill repertoire building. Continuous Recording and Trial by Trial Recording.

Continuous Recording:

Continuous Recording involves recording every instance of behavior. This means observing, measuring and recording the target skill each and every time it occurs. Let’s look at each component of the continuous recording method in more detail:

  • The Observation Period: Which is the specific time frame when the behavior is likely to occur. During which the therapist should be prepared to observe and measure the behavior. Typically with the continuous recording method the observation period is an entire session. The observation period may also be a shorter period of time such as the first hour of a session or during a particular activity.

Example I:

A child is learning to ask for help. If the observation period is an entire session than each instance of asking for help (the target behavior)  would be observed and recorded.

Example II:

A child is learning to ask for help during art activities. If the observation period is the during a specific activity  than each occurrence during an art related activity would be observed and recorded.

  • When to Record Data: In continuous recording the data is recorded every time a behavior occurs. This means that the therapist records the measured behavior immediately following each occurrence of the target skill. Because each and every instance of behavior is recorded within continuous recording it is possible to measure and record several dimensions of behavior, including frequency, duration and latency.

Types of measurement used in Continuous Recording

  • Frequency: The frequency of behavior is the number of times the behavior is observed. To measure the frequency of behavior the therapist counts how many times the behavior occurs during the observation period. For example, to measure the frequency of a child’s mands, the therapist “makes a tally mark” or “clicks the appropriate button” Likewise, to measure the frequency of a child’s eye contact, the therapist would also tally or click the applicable skill behavior. The total number of target behaviors observed is the frequency of a given behavior.
  • Duration: The duration of behavior is the total amount of time a child is engaged in the behavior, from start to finish. In other words the duration is how long the behavior lasted. To measure the duration of behavior the therapist times the behavior from its onset to its conclusion and records to elapsed time (in the applicable area under the applicable behavior). For example, to measure how long a chid plays with their peers the therapist would start the timer when the child began playing near their friends and stop the timer when the child moves away from their friends. This same method of measurement would be used to record the duration of a conversation, or how long it takes a child to get dressed. The entire time the child engages in the specific behavior is the duration.
  • Latency: Latency is the amount of time between a stimulus event and the onset of behavior. In other words the latency is how long it takes to start a behavior after a stimulus has been presented. To measure the latency of behavior, the therapist starts the timer upon presentation of the SD and stops the timer once the child engages in the target behavior. The total time between the stimulus event and the onset of behavior is the latency.

Trial by Trial Recording:

Trial by trial recording is recording a child’s response to a discriminative stimulus on a Trial by Trial basis. Practically speaking, trial by trial recording includes observing, measuring, and recording a child’s response following each trial. Let’s look at each component of the trial by trial recording method in more detail.

  • The Observation Period: Which is the specific time frame when the behavior is likely to occur. During which the therapist should be prepared to observe and measure the behavior. Typically with the trial by trial recording method the observation period is during the lesson targeting the specific skill to be observed.

Example I:

A child is learning to follow instructions. The therapist will use the trial by trial recording method to collect data on the new skill of following instructions. The observation period is during the child’s following instructions lesson. So the therapist must be prepared to observe, measure and record every time the child follows instructions after each trial during the following instructions lesson.

Example II:

A child is learning to imitate sounds so the observation period is during the echoic lesson. And the therapist must be prepared to observe, measure and record every time the child imitates sounds following each trial during the echoic instructions lesson.

In trial by trial recording the data is recorded at the end of every trial, this means that the therapist records the behavior during each intertrial interval that occurs between trials. Recall that the intertrial interval is the period of time between the end of one trial and the start of the next. So during trial by trial recording would record data after the current trial but before the next trial.

Example I:

A child is learning to follow instructions. In the first trial the therapist presents the SD “sit down”. The child responds by sitting down, and the therapist gives the child a piece of cookie. The therapist than records the child’s response before beginning the next trial of the following instructions lesson.

Let’s look more closely at how the child’s response is recorded in trial by trial recording. Recording a child’s response to an SD includes observing, measuring, and recording the accuracy and independence of the child’s response. As previously discussed, there are four types of responses in discrete trial instruction:

  1. Correct – The correct response is the target response specified for a given Sd.
  2. Incorrect – Any response that is different from the target response.
  3. No Response – When a child does not react at all to the Sd.
  4. Prompted – When the therapist provides a prompt in addition to the Sd to increase the probability of a correct response.

Example I:

A therapist presents the SD “clap”, the child claps, the therapist provides a reinforcer and the therapist records the response as correct.

Example II:

A therapist presents the SD “clap”, the child waves, the therapist applies the appropriate correction procedures and does not provide a reinforcer and the therapist records the response as incorrect.

Example III:

A therapist presents the SD “clap”, the child does not respond, the therapist applies the appropriate correction procedures and does not provide a reinforcer and the therapist records the response as no response.

Example IV:

A therapist presents the SD “clap” with a modeled prompt, the child claps, the therapist provides a reinforcer and the therapist records the response as prompted response.

So far, you have learned about two recording methods: continuous recording and trial by trial recording. And the various types of measurement used to record behavior. The written record of these measures is called raw data. In other words, the frequency, duration and latency of each behavior recorded with the continuous recording method and the Correct, Incorrect, Prompted and No Response measures of the Trial by Trial method of recording are the raw data. For example, the “tally” marks recording the frequency of mands is raw data, the minutes and seconds recording duration of dressing are raw data, recording a response as correct, incorrect, etc. is raw data. This information is critical, but it can be difficult to see a child’s overall skill performance simply by looking at the raw data. As a result, raw data is often converted to summary measures.

Three Types of Summary Measures:

Summary measures “summarize” a child’s performance on a particular skill. There are three summary measures covered in this lesson:

  • Rate – The frequency of a behavior per time. Practically speaking, rate is the number of times the behavior occurs in a specified amount of time. Rate is calculated when frequency is divided by a specified amount of time. The time is specified by the BCBA , however in many cases the time is one hour. Therefore the rate is the frequency per hour (like miles per hour while driving). For example, to measure the frequency of a child’s mands, the therapist would record the number of mands and the BCBA specified that the rate should be frequency per hour than the therapist would divide that number of mands by the time.
  • Percentage – Percentage is used to compare how one measure of behavior compares to another measure of behavior. In other words, percentage is a comparison of two quantities or types of data. Percentage is calculated when one measure of behavior (Behavior “A”) is divided by another behavior measure (Behavior “B”), the resulting proportion is than multiple by 100% to get a percentage measure.
    • Percent of opportunities: Percent of opportunities is the number of times a behavior is observed  compared to the number of times the child had the opportunity to engage in the behavior multiplied by 100%. For example to measure the percent of opportunities that a child asked for help the therapist divides the number of times the child asked for help with the number of time she needed help and multiplied that by 100% to gain the percentage of opportunities taken.
    • Percent Correct: Percent correct is the number of correct responses compared to the number of times a response is provided. (And works the same way as percent of opportunities)
    • Percent Behavior Chains: When teaching a complex skill using chaining the percentage of steps in the behavior chain completed independently is common summary measure. The percentage is calculated by dividing the number of steps completed independently by the total number of steps in the chain and multiplying that proportion by 100% to obtain the percentage rate.
    • Percent of Time: Percent of time is the duration of the target behavior compared to the duration of the observation period. In other words, the duration of each episode is added together to achieve the proportion of total target behavior duration. This is then divided by the total observation duration and multiplied by 100%. For example, to measure the amount of time a child plays with their peers, the therapist divides the amount of time the child spent playing with peers by the total observation period and multiplies that proportion by 100%
  • Average – An average is a value that represents a larger set of values. In other words, an average is a summary measure that describes a typical level of behavior observed (think batting average).
    • Average Duration: Average duration is how long the behavior typically lasts. Average is calculated by calculating the total target behavior duration by the frequency of the target behavior. For example, if a child has 5 conversations for a total duration of 30 minutes the therapist divides the total duration (30 minutes) by the number of conversations (5) to obtain the average (6 minutes per conversation).
    • Average Latency: Average Latency is how long it typically takes to start the behavior after a stimulus has been presented. Average latency is calculated by dividing the total latency by the number of responses made. For example, the therapist asks a child to clean up 4 times during a session. The child starts cleaning after 15 seconds, 1 minute, 45 seconds and 2 minutes respectively. The therapist adds these proportions together and then divides that by the number of opportunities to respond (4 times asked to clean) to calculate the average latency.

This lesson has discussed the terms and procedures used for data collection in skill repertoire building. It is essential to be familiar with the data recording methods, the types of measures used to record raw data, and how data is summarized. A therapist will be required to collect data on multiple skills throughout every therapy session. This data must be reliable as the BCBA will use this data to make critical decisions regarding, what, when, and how to teach each child based on the data.


The Discrete Trial Data Sheet

In the last lesson, you learned about the terms and procedures used in data collection for Skill Repertoire Building including two recording methods: continuous and trial by trial. This lesson will describe how to complete a discrete trial data sheet, one of the forms used to collect data in trial by trial recording. A Discrete Trial Data Sheet is one of the forms used by the therapist to record complete and accurate data on a child’s responses during discrete trial training. A Discrete Trial Data Sheet can vary, however the required collected information remains similar. There are three components to a Discrete Trial Data Sheet: 1) Identifying Information, 2) Description of the Current Lesson, and 3) Trial by Trial data.

The Identifying Information for Discrete Trial Data Sheets

The Identifying Information section includes the full name of the child whose data is being recorded (though in some cases this may be abbreviated according to the necessities of discretion and security). The title of the lesson being taught, and the month/year that data was recorded. The identifying information must be completed each time that a new data sheet is used.

The Lesson Description section includes:

  • The provider of the instruction: The provider is the therapist who is teaching the target skill to the child. In this area the therapist records their initials so it is clear who recorded the data.
  • The date the lesson was practiced: to include month/day/year
  • The current Sd: The Sd or Discriminative Stimulus must be written in full for the current lesson and must clearly describe any vocal or visual stimuli presented
  • The target response: The response expected from the child
  • The Discrimination Training Procedure used (MT, ET, or RR)
  • Acquisition Target: As previously learned in Massed Trials (MT) the Sd is presented repeatedly across consecutive trials, so when conducting Mass Trials the data is collected on a single Target. This data is recorded in the section labeled ‘T1’ as in Target 1. However, during Expanded Trials and Random Rotation, two or more Sds are presented across consecutive trials, and this means that the data is collected on multiple targets. So the therapist would indicate in this section whether or not the data was collected on a “D” distractor trial, or a “T2” second target.

Example:

A therapist is using Discrete Trial Training to teach a child to imitate non-verbal actions. On the data sheet the therapist would record their initials, the date of the lesson and the discrimination trial used in the lesson, whether Massed Trials, Expanded Trials, or Randomly Rotated. In the Sd section of the sheet, the therapist would record what they would say and do to teach the child, in other words “Do this” + model (action). Likewise they would record the expected response which would look like, child imitates (action). In this example the child is being taught using Random Rotation using Distractor trials with known distractors. Therefore, the therapist would circle “D” for distractor trials and indicate that the distractor is previously “Mastered Targets”.

The Trial by Trial Data Section is used to record the specific trial by trial data. This includes:

  • The Stimulus Presented: In this area, the therapist records any stimulus presented not already specified on the Sd portion of the form, including vocal and visual stimuli. For example, in using the same example of teaching a child non-vocal imitation, the therapist already recorded the SD “Do This”, however because the trials are random rotation, the SD presented changes. So here the therapist records the specific action modeled for each trial, i.e., clap, wave, etc.
  • The Type of Response: Correct, Incorrect, No Response, or Prompted Response. Here the first column is used for trials of the first target, the second column is used for distractor or additional targets.
  • The Actual/Specific Response: Incorrect responses are recorded here noting specifically how the child actually responded to the SD. I.e., clapping instead of waving for the SD, “Do this” <wave>.
  • The Prompt Used: When a child’s response is prompted the specific prompt(s) used must be recorded here.
  • Comments: Any additional important information would be listed here. Such as, type of distractors used, effective prompting techniques, inappropriate behaviors observed, or instructions to other therapists regarding which step of discrimination training procedure to use during the next session
  • Summary Data: The raw data is summarized as percent correct. Percent Correct Calculations include:
    • Target 1 Responses: This provides the BCBA with a summary measure of correct responses to Target 1, or how often the child responds correctly to the target. This does not include distractor or target 2 responses. This is calculated by dividing the total target 1 correct responses by the number of target 1 trials given and than multiplying that proportion by 100%.
    • Distractor or Target 2 Responses. This works exactly the same way as target 1 responses except that this includes D or T2 data.
    • Overall Responses. The Overall Percent Correct Responses provides the BCBA with a Summary measure of all correct responses combined. So here all correct responses are divided by the total number of trials presented and than multiplied by 100%.

As previously learned, during trial by trial recording, data is recorded during each inter-trial interval. Therefore after each trial, there therapist records the stimulus presented, the response type, the actual response, and the prompt used if applicable. At the completion of the lesson, the therapist summarizes the raw data by calculating the percent correct.

This lesson has introduced you to the three sections of the Discrete Trial Data Sheet, and explained how to record trial by trial data. As a therapist it is important to accurately and completely record the necessary Identifying Information, the Lesson Description Information and the Trial by Trial Data as well as the Percent Correct Summary Measures. The BCBA uses this data to make decisions regarding the learning program for each child and therefore it is critical that this information is reliable.


The Task Analysis Data Sheet

In the last lesson you learned how to complete a discrete trial data sheet one of the forms used to collect and record data in trial by trial recording. This lesson will explain how to complete a task analysis data sheet, another form used to collect and record data in trial by trial recording. The Task Analysis Data Sheet is used to record data when teaching behavior chains ( A behavior chain is a complex behavior comprised of multiple single SD-Response components that occur in a specific sequence). Therefore data collection for a behavior chain involves observing and recording data on each SD-Response component of the behavior chain that the child is required to perform during each learning trial. The data is collected and recorded on a child’s accuracy and independence when performing each step in the behavior chain.

When teaching a behavior chain the BCBA may require that a Task Analysis Data Sheet is used to record complete and accurate data on the child’s responses. The Task Analysis Data Sheet has Four Sections: Identifying Information, Description of the Current Lesson, Specific Trial Data and Summary Measure Information. As in other lessons, the sample form displayed in this lesson is simply a guideline and while the data will remain consistent the format and structure may vary.

Identifying Information Recorded on the Task Analysis Data Sheet:

The Identifying Information section includes the full name of the child whose data is being recorded (though in some cases this may be abbreviated according to the necessities of discretion and security). The title of the lesson being taught, and the month/year that data was recorded. The identifying information must be completed each time that a new data sheet is used.

The Lesson Description section includes:

  • The Target Behavior: The Target Behavior for the lesson must be written in full and clearly describe the behavior chain being taught.
  • The Initiating Stimulus: The initiating stimulus for the behavior chain must be recorded, recall that the Initiating Stimulus is the Stimulus Condition or SD that initiates the behavior chain.
  • The Chaining Procedure Implemented: The specific chaining method being used to teach the skill must be listed, in other words, whether it is Forward Chaining, Backward Chaining, or Total Task Presentation.
  • Provider: This is the therapist who is teaching the target skill. The therapist inputs their initials into this section so it is clear who recorded the data.
  • Date Practiced: mm/dd/yy format
  • Steps: Here each step or component of the chain will be listed. The BCBA identifies the specific steps of the behavior chain by conducting a Task Analysis. A Task Analysis being the process of breaking down a behavior into a sequence of SD-Response components. A task analysis is individualized for each child, and lists the sequence of steps the child must perform to complete the behavior chain. Each of these SD Response components of the chain is recorded by the therapist in the Steps section of the Task Analysis Data Sheet.

The Specific Trial Data Section is used to record the specific trial data. In other words the child’s accuracy and independence for each step or SD/Response component of the behavior chain. This includes:

  • Time Observed: The time observed is the exact time that the behavior chain began and the time that the behavior chain was completed.
  • Step-by-Step Data: In this section the type of response to the given step in the chain is recorded either as (+) Correct, (-) Incorrect, (NR) No Response, etc.  There is a scoring key on each TADS form to record the specific response:

The key identifies the notation to use for each type of response. If a child responds correctly to a step

The Summary Measure Information Section is used to summarize the raw data:

  • Percentage Completed Independently: Percentage completed independently is the number of correct responses divided by the total number of steps in the chain multiplied by 100%.
  • Total Duration: The duration of behavior is the total amount of time engaged in the behavior from start to finish. In teaching behavior chains the entire sequence is considered One Complex Behavior. So the total duration refers to the time spent performing the entire chain. This is calculated by subtracting the stop time from the start time recorded in the time observed section.

Chaining Procedures and the Task Analysis Data Sheet

Forward Chaining: In forward chaining one SD-Response component is taught at a time. And then chaining the components together in the specified sequence starting with the initial SD response component and continuing forward through the remaining components. In forward chaining, the BCBA may specify that the lesson stop once the child completes the target behavior for a specific component of the behavior chain. In other cases the BCBA may instruct the therapist to complete the chain for child in a given trial. In either of these cases data is only collected on the steps that the child is currently learning or has previously mastered. Data is not collected on steps NOT performed or practiced on by the child.

Backward Chaining: In backward chaining the therapist starts with the final SD-Repsonse component and works backward. Again, only a single SD-Response component is taught at a time. In backward training the therapist may be instructed to produce the Final SD for the last step in the chain or perform the initial step in the chain leading up to the final SD/Response component for the child. Just as in forward chaining, data is only collected on steps the child is currently learning or has previously mastered. Data is not collected on steps NOT performed or practiced on by the child.

Total Task Presentation: In total task presentation all steps of the Sd-Repsonse chain are taught during every learning trial. So the child is prompted through the total task or each step of the behavioral chain on every learning trial. As a result for total task presentation data is collected and recorded for all steps. When steps are performed incorrectly the therapist divides the data box and records a minus for the incorrect response and then indicates which prompt was used to help the child in the error correction procedure.

This lesson has introduced you to the three sections of the Task Analysis Data Sheet, and explained how to record task analysis data. As a therapist it is important to accurately and completely record the necessary Identifying Information, the Lesson Description Information and the Trial Data as well as the Summary Measures. The BCBA uses this data to make decisions regarding the learning program for each child and therefore it is critical that this information is reliable.


Conducting Probes

As you have learned children with Autism have many skill deficits in a variety of developmental domains, such as language, play, academic skills, executive functions, motor skills, social skills, self-help, perspective taking, and while some of these skill deficits may be obvious, others may not be. Often times a child’s caregivers and therapy team are uncertain about whether or not a child knows a particular skill. Any time there is uncertainty about whether or not a child knows a skill it is important that child’s ability to perform that skill is assessed to avoid spending valuable time teaching an already known skill.

In this lesson you will learn how to conduct probes to assess a child’s ability. The specific data collection requirements for recording a probe will be covered in another lesson.

What is a Probe?

A probe is a method of assessing a child’s skills to determine what a child knows. In other words, a probe is a test of what the child can and can not do. This means the child is given an opportunity to behave in a particular way. Their response is observed and recorded and the data is used to determine if the child can perform the skill correctly and independently. By conducting probes the BCBA can determine exactly what target to introduce and the child’s therapy team does not waste time teaching skills the child already knows. Probes also allow the BCBA to determine when a skill has been learned; that is, a child can perform a recently taught skill with both accuracy and independence. There are two types of probes: Pre-instructional and Post-instructional.

Two Types of Probes:

Pre-Instructional

A pre-instuctional probe or a baseline probe is conducted prior to teaching a skill in order to determine which skills need to be taught, in other words what they know and can perform and what they do not know and cannot perform.

Example:

A BCBA is unsure that a child knows their colors. In order to determine the child’s ability the BCBA requests that a pre-instructional probe be conducted to determine what colors if any the child knows. During the baseline probes the child was unable to identify the color blue, so the BCBA determined the color blue to be the first target in their color labelling lesson.

Post-Instructional

A post-instructional probe or a mastery probe is conducted concluding the teaching of a new skill to ensure that child can perform the recently taught skill accurately and independently or if the skill requires additional practice. Mastery probes are typically conducted following several sessions that have indicated that the child is performing the skill accurately and independently. The probes ensure that the skill has been mastered and moreover has generalized across several therapists and maintains in the absence of reinforcement.

Example:

After teaching the child the color blue, the child begins to accurately and independently identify the color blue in several sessions of random rotation with known distractor items. And the data suggests that the child has mastered this skill. So the BCBA requests that the therapist conducts a post-instructional probe to confirm that the child has indeed learned the skill or requires additional instruction.

Note that pre and post instructional probes differ only in when they are conducted, either before teaching a new skill or after teaching a new skill.

Before learning how a probe is conducted it is important to understand the assessment environment. In an ideal situation an assessment of a child’s skills would take place in multiple settings such as home, school, and their natural environment using a variety of stimuli and be conducted by a number of different individuals over multiple sessions. This would ensure that the child knows the skill, and has generalized that skill across environmental influences (places, situations and people). However, such extensive assessment procedures would take substantial time and effort to implement and could delay further instruction. Probes are designed to adequately address the natural environment needs, ensure the child’s ability while avoiding a delay in instruction.

In the natural environment a child must respond accurately and independently to a variety of people at different times with different distracters. To adequately address these needs, a probe is typically comprised of 3-5 opportunities to respond and is conducted by multiple therapists, during multiple therapy sessions in a random rotation format. With multiple therapists conducting probes the BCBA can ensure that the skill is generalized to a variety of people and is not limited to a particular therapist. By conducting probes over multiple sessions the BCBA can ensure that the skill has been maintained over a short period of time since the probe trials are separated by a few hours or a few days. And while some trials are conducted over a series of days, others must be conducted over a single day. However, those trials are not conducted by a single therapist but rather requires a minimum of three different individuals to conduct the probe. Similarly, the BCBA uses random rotation to ensure that child has generalized this information across distracters.  Each probe is followed and proceeded by several distractors. The probes may include known and unknown distracter items and may also include probes for different skills.

How is a Probe Conducted?

A probe consists of two parts: The SD and the Response. In the natural environment a reinforcer is not provided for every single response a child makes and since a probe is meant to be a test of a child’s skills and is used to determine how a child may respond in the natural environment no consequences are provided during a probe. As previously discussed a consequence is the outcome that immediately follows the child’s response. When teaching a child a new skill there are two possible consequences that may follow a response: Reinforcement to increase correct responding and an Error Correction Procedures to decrease incorrect responding. Consequences are critical when teaching  a child a new skill. Without consequences no learning would take place. However, when conducting probes the therapist is not teaching a new skill. Instead, the therapist is determining if the child already knows a particular skill. This means that when conducting probes therapists do not provide any consequences that may increase or decrease the child’s responding. This is an important difference between teaching and conducting a probe for a skill. This means that the therapist does not provide an SR (Reinforcer) following a correct response nor do they use an Error Correction procedure following an incorrect response. During a probe, the therapist simply observes and records the child’s response.

Lastly, it is important to note that because reinforcers are not provided for correct responses during probe trials the child may quickly lose motivation to continue responding. So how do you motivate a child to continue to respond without providing a correct respond during a probe. First, an SR (Reinforcer) can be provided for a correct response to a distractor stimuli presented during random rotation. Another option is to provide reinforcers for appropriate behaviors the child engages in during a probe (for example, the therapist could praise the child for sitting nicely, or tickle the child when they make eye contact. Using these techniques allows the therapist to conduct the probes correctly while maintaining the child’s motivation/interest.

Example:

A child has mastered blue, yellow and red in their color identifying lesson. In other words the child has been consistently and accurately able to identify these three colors during lessons when these colors are presented. The BCBA has requested that a pre-instructional probe be conducted to see if the child also knows orange, purple and green or if they need to be introduced as targets in additional color identification lessons.

So on Monday Therapist 1 is working with the child and places a purple, yellow and green card in front of the chid and presents the SD, “touch green”. The child responds correctly and touches the green card. The therapist than changes the stimuli on the table and presents green, purple and blue cards and says, “touch purple”. This time the child does not respond, the therapist records this and moves on to the next trial. Next is presented blue, yellow and purple in a distracter trial and the therapist says, “touch yellow”. The child touches yellow and the therapist reinforces the correct response with social praise. Lastly, the stimuli are changed again and the child is presented with purple, orange, and blue cards along with the SD “touch orange”. The child touches the purple card and the therapist records this as an incorrect response.

So on Tuesday Therapist 2 is working with the child and conducts the second probe trial. The therapist places a purple, blue and orange card in front of the chid and presents the SD, “touch purple”. The child touches the orange disk and the therapist records this as an incorrect response. Next the therapist presents a distracter trial and presents a blue, orange and purple card along with the SD, “touch blue”. The child responds correctly and the therapist provides a reinforcer for the correct response to the distracter item. The stimuli are then changed and the therapist conducts another distracter trial and presents to a purple, orange and red card along with the SD “touch red”. The child again responds correctly and the therapist provides social praise for the correct response on the distractor trial. Next the therapist presents a purple, green and orange card along with the SD, “touch orange”. The child touches the purple card and the therapist records this response as incorrect. The stimuli are changed on last time and a green, orange and purple card are place on the table along with the SD, “touch green” the child correctly identifies the green and the therapist records this response.

So on Tuesday Therapist 3 is working with the child and conducts the third probe trial. The therapist places a purple, green and orange card in front of the chid and presents the SD, “touch green”. The child correctly identifies the green and the therapist records the response. While the therapist prepares the SD for the next probe trial the child waits nicely in their seat so the therapist uses that opportunity to provide social praise for the child’s appropriate behavior. Then the next set of stimuli is presented, purple, orange, and green cards along with the SD, “touch purple”. The child touches the orange and the therapist records the incorrect response. Again, the stimuli are changed to purple, orange and yellow and presents the SD “touch yellow” as a distracter trial. The child responds correctly and is provided social praise. Finally the stimuli is presented as green, purple and orange cards along with the SD “touch orange”. The child touches the purple and the therapist records this as an incorrect response.

When the BCBA receives the results of the probe they determine that the child already knows the color green and introduces only purple and orange in the color identifying lessons as new targets.

This lesson has discussed the terms and procedures used in conducting probes. As a therapist it is essential to know when to conduct the two types of probes, and the difference between probe and instructional procedures.  A therapist will be required to conduct probes on a regular basis. The program supervisor uses the data from probes to determine what skills a child needs to be taught versus the skills a child already knows. The ability for a therapist to conduct probes accurately is critical to a child’s success.


The Discrete Trial Probe Data Sheet

This lesson introduces you to the Discrete Trial Probe Data Sheet and described how to collect and record complete, accurate data when conducting probes. Recall that a probe is a methods of assessing a child’s skill’s to determine what a child knows. In other words, a probe is a test of what the child can and cannot do. The Discrete Trial Probe Data Sheet is used when determining whether or not a child is able to perform a skill. Recall that data is collected before, during and after teaching a new skill in order to identify skill deficits, monitor progress, determine mastery and ensure maintenance and generalization of a skill. The Discrete Trial Probe Data Sheet is used to identify skill deficits by collecting data before and after teaching a skill. It is used before teaching a skill to determine which skills need to be taught and which skills a child already knows and can perform independently. It is used after a skill has been taught in order to  ensure that mastery including maintenance and generalization of a skill is achieved.

When is the Discrete Trial Data Sheet Used:

A Discrete Trial Probe Data Sheet allows the therapist to record complete and accurate data on a child’s responses during pre-instructional (baseline) probes and post-instructional (master) probes. The discrete trial probe data sheet used in the lesson (like all other sheets, is variable between BCBA’s and agencies, but the information within will remain consistent between sheets). The Discrete Trial Data Sheet has four basic sections:

The Identifying Information section includes the full name of the child whose data is being recorded (though in some cases this may be abbreviated according to the necessities of discretion and security). The title of the lesson being taught, and the month/year that data was recorded. The identifying information must be completed each time that a new data sheet is used.

The Lesson Information section includes the discriminative stimulus SD, and the expected Response. The SD must be written in full for the current lesson and clearly describe any vocal or visual SDs presented. The response expected from the child is also described.

Example:

In a child’s object matching lessons the BCBA has requested that the therapist conduct a probe to test the child’s object matching skills: So the description for the SD would be “Put with same” while giving an object. The response would be the child matching (object).

The Specific Probe Trial Information section is used to record probe trial data. This section includes the following for each probe trial:

  • The Stimulus Presented: In this area the therapist records specific details not already included on the SD portion of the form (in other words the SD may be “give me” or “match” or “touch” but the object to be given, matched, or touch may change and this should be included here. This will include any not previously mentioned vocal or visual stimuli presented to the child. It is important to include the stimuli presented because this information specifies the SD-Response component that is targeted for probing.
  • The Type of Response: Correct/Incorrect
  • Date Conducted/Initials of Therapist: mm/dd/yy and initials

Example 1:

The BCBA has instructed the therapist to conduct a baseline probe for the objects shoe, ball and block. When conducting baseline probe trials for the shoe, the therapist presents the SD “put with same” while giving the child a shoe. So the therapist writes ‘shoe’ on the first line of the data sheet in the Stimulus Presented column indicating that matching a shoe is a probe target. The vocal SD “put with same” is not written because that information is included in the lesson section of the data sheet and remains the same for every probe trial in this particular lesson.

The Discrete Trial Probe Data Sheet has 5 Columns for recording probe trial data: P1, P2, P3, P4 and P5. This is in keeping with a probe typically being comprised of 3 to 5 opportunities to engage in the behavior and determine if a skill is known. The exact number of probe trials that should be conducted will be determined by the child’s responses. A minimum of 3 probe trials for each target must conducted by 3 different therapists. Data is collected and recorded for each of these three trials in the applicable columns label. If the child responds correctly on all three of the probe trials additional trials are not needed and therefore any remaining columns would remain unused. If a child responds incorrectly on one or two of the three trials, 5 trials are needed. However three or more incorrect responses negate additional probes as the child has not mastered the lesson.

Recall that when conducting probe trials that the therapist does not provide a reinforcer following a correct response. Instead the therapist simply observes and records the child’s responses. Additionally, the therapist does not engage an error correction procedure for incorrect responses. For each probe trial the therapist must additionally record the date the probe was conducted and initial the entry so that it is clear who conducted the probe trial and recorded the data.

A completed specific trial section of a Discrete Trial Probe Data Sheet would look like the following:

The Summary Measure and Outcome Information section is where the raw data is summarized as percent correct and the outcome of the probes is indicated.

Following the final probe trial for a specific target response the therapist must convert the raw data to percent correct. To obtain the percent correct, the therapist takes the number of correct responses for each target and divides that by the total number of trials for that target than multiples the proportion by 100%

Finally, the therapist records the outcome of the probe. The outcome of the probe is the overall description of the child’s ability to perform the skill independently and correctly. The Outcome Section of the Outcome Information has two columns: Pre-Instruct (Baseline)and Post-Instruct. (Mastery).

In Pre-Instruction Probes when a child performs a skill correctly 80-100% of the time than that child is considered to know the skill and the “K” is circled. When a child responds correctly less than that than the skill is consider to be “U” unknown. Likewise, in a Post-Instructional probe when a child responds accurately and independently  80% or greater than the “M” is circled as the child is considered to have mastered that target. Likewise, if the percent correct is less than 80% than the RE-1 is circled and that target must be practiced.

This lesson has introduced you to the four sections of the Discrete Trial Probe Data Sheet and described how to record data for baseline and mastery probes. As previously discussed it is important that the therapist record the necessary identifying information, a description of the current lesson, specific data for each probe trial and summarized that data in the Outcome section. The BCBA uses this information to determine what skills are learned and what skills still need to be taught. So recording accurate data ensures the child’s lessons and targets are appropriate.


The Task Analysis Probe Data Sheet

This lesson introduces you to the Task Analysis Probe Data Sheet one of the forms used in trial by trial recording. The Task Analysis Probe Data Sheet is used to collect data when determining if a child knows a complex skill. Recall that data is collected before, during and after teaching a new skill in order to identify skill deficits, monitor progress, determine mastery and whether or not the skill has been generalized across environments and is maintained in the absence of a reinforcer. The Task Analysis Probe Data Sheet is used to identify whether a child can accurately perform each step in the behavior chain. And while format and structure will vary among BCBA’s and agencies, the essential information contained within will remain the same.

The Task Analysis Probe Data Sheet (TAPDS) is used when determining whether a child can perform a complex skill or behavior chain comprised of several steps. Recall that data collection for a behavior chain involves observing and recording data on each SD-Response component of the chain. A TAPDS allows the therapist to record complete and accurate data on a child’s responses during pre-instructional and post-instructional probes of the behavior chain. Data recorded on the TAPDS allows the BCBA to determine whether or not the child is able to perform the entire data chain as well as each of the specific steps in the behavior chain. The TAPDS has 4 sections:

The Identifying Information section includes the full name of the child whose data is being recorded (though in some cases this may be abbreviated according to the necessities of discretion and security). The title of the lesson being taught, and the month/year that data was recorded. The identifying information must be completed each time that a new data sheet is used.

The Lesson Information section includes:

  • Target Behavior: The target behavior must be written in full and clearly describe the behavior chain being probed.
  • Initiating stimulus SD: The initiating stimulus must clearly be described. Recall, the initiating stimulus is the stimulus condition or SD that initiates the behavior chain.
  • Therapist: Is the person who conducted the probe. Here once again the therapist records their initials so that it is clear who recorded the data.
  • Date Probe was Conducted: mm/dd/yy
  • Steps: Here the therapist records each step (SD-Response component) in the behavior chain.

Example:

The BCBA requests that a baseline probe is conducted to determine if a child can put on their shoes when instructed. Therapist 1 starts the TAPDS form, by filling in the target behavior as , Putting on Shoes, writing the initiating stimulus as “Put on Shoes” and then listing the steps of the behavior chain as perviously determined by the BCBA during a task analysis.

The Step by Step Probe Trial Information section is used to record probe trial data. This section includes the following for each probe trial:

Here the therapist will record the accuracy of each step or each SD-Response component of the behavior chain. Data on the child’s response is recorded in the appropriate data column (in other words Probe 1 is recorded in column P1, Probe 2 is recorded in column P2, etc.). As previously discussed, there are only two types of responses when conducting probes, and that is either correct or incorrect. As the scoring key indicates if the child responds correctly the appropriate step and column are marked with a “+”, if they respond incorrectly once again the applicable step and column are marked with a “-“.

Remember that when conducting probes the therapist is not attempting to teach the child a new skill. With behavior chain probes the therapist is determining if a child already knows each step of the chain individually and if the child can perform that entire chain accurately and independently. Recall that during probe trials consequences are not provided as they may increase or decrease the child’s response. This means that the therapist does not provide reinforcers for correct responses or error correction procedures for incorrect responses. The therapist simply observes and records the the child’s response on the data sheet.

As previously discussed a behavior chain is a sequence of SD/Response components, where each response in the sequence produces the SD for the next response in the sequence (rather one thing leads to another). This means that when conducting a probe for a behavior chain if a child cannot independently perform a step in addition to recording a minus sign for the incorrect response the therapist must complete that step for the child in order to initiate the next SD in the chain.

A probe is typically comprised of 3 to 5 opportunities for the child to engage in the target behavior(s) in order to determine if the skill is known. The exact number of probe trials is determined by the child’s responses. In behavior chain probes the entire chain is one trial, comprised of multiple steps. When conducting behavior chain probes at least 3 trials for each target behavior are conducted and by multiple therapists. The step-by-step data is recorded for each of these trials in the step by step data section of the TAPDS, in the applicable row (step in the chain) and column (which probe trial is being conducted, i.e., Probe 1 = P1, Probe 2 = P2, etc). If a child performs all of the steps correctly across the first three trials no more trials are needed and the child has mastered the behavior chain or already knows the skill. Alternately if a child gets 1 to 2 steps wrong within 1 trial  then an additional 2 trials will be needed to determine whether or not the child knows each step of the behavior chain and can perform the entire chain accurately and independently. However, if a child gets one or more steps wrong across two or more trials than no further trials are needed because the child does not know the steps of the chain and cannot perform it accurately and independently.

A completed lesson and step-by-step section would look like the following:

The Summary Measure and Outcome Information section is where the raw data is summarized as percent correct and the outcome of the probes is indicated.

Following the final probe trial for the target behavior the raw data is summarized as percent correct. Percent correct is calculated by taking the number of correct responses and dividing it by the total number of responses for that particular step and then multiplying the resulting proportion by 100%. With behavior chain probes the percent correct is calculated in two ways, first for each probe in the behavior chain and then for the each step across probe trials.

In using the above example for putting on shoes the raw data recorded would look like the following when summarized:

In looking at Outcomes, the outcome of the probe is the overall description of the child’s ability to perform the skill independently and correctly. There are two columns within the outcome section of the data sheet, one is for Pre-instructional probes, the other for Post-instructional probes which means that this sheet can be used for both baseline probes and mastery probes.

In behavior chain probes two outcomes are determined.

  1. Each step in the behavior chain across trials
  2. Overall outcome of the target behavior probe trials

As previously learned in other data sheets, in a pre-instructional probe an accuracy rate of 80-100% indicates that the target is known = K. Whereas,79% or less indicates that the target is not known = U. Similarly with post-instructional probes, an accuracy rate of 90-100% indicates that the target skill is mastered = M, and an accuracy rate of 79% or less indicates that the skill has not been mastered and must be reintroduced for continued practice = RE-I.

The second outcome that is determined is the overall outcome of the target behavior probe trials. When conducting a pre-instructional probe, when all of the steps are known than the skill is considered to be known = K and is circled in the overall outcome section. However, if any of the steps in the behavior chain are unknown = U, then U is circled in the overall outcome section. The same method applies to post-instructional probe outcomes. If all of the steps are known then the target is considered mastered = M, and the M is circled. Additionally, if any of the steps are unknown =RE-I, than the RE-I is circled and the chain is not mastered and must be reintroduced for continued practice.

This lesson has introduced you to the four sections of the task analysis data sheet and described how to record data for behavior chain probes. As previously discussed it is important that the therapist record the necessary identifying information, a description of the current lesson, specific data for each step in the behavior chain and summarized that data in the Outcome section. The BCBA uses this information to determine what complex skills and specific targets are learned and what skills still need to be taught. So recording accurate data ensures the child’s lessons and targets are appropriate.

Learn More…

Behavior Management Data Collection

Leave a Reply

Your email address will not be published. Required fields are marked *