Master ABA

Data Collection Methods: Continuous vs Discontinuous Measurement in ABA

The field of Applied Behavior Analysis (ABA) relies heavily on data to make informed treatment decisions. Professionals in the field must choose the data collection method that measures the right behavior. They analyze data to determine the effectiveness of interventions. If the data demonstrate progress, interventions continue. If the data reveal a trend in the wrong direction, the professional changes the intervention.

The right data collection system provides information needed to determine the effectiveness of programming. Choose either a continuous data collection method (frequency, rate, duration, or latency) or a discontinuous data collection method (partial interval, whole interval, or momentary time sampling). The right method provides accurate data that are sensitive to behavior change. Continuous methods provide the most accurate data, although they can be difficult to utilize in busy environments. Discontinuous methods offer an estimate of the occurrence of behavior, but can be used even when staff can’t attend to the learner’s behavior throughout the entire session.

Whether you’re an RBT or BCBA candidate studying for the exam or a new professional just starting out, these data collection methods can be confusing.

Download this infographic to help you keep this information fresh!

Continuous and discontinuous data collection methods in ABA infographic, frequency/rate, duration, latency, partial interval, whole interval, momentary time sampling

Contents

Continuous Data Collection Discontinuous Data Collection Other Data Collection Methods Advantages and Disadvantages of Each Type of Data Collection Method 5 Ways to Make Frequency Data Collection Easier and More Discrete Research in Data Collection Methods Ethical Considerations When Choosing Data Collection Methods Research Related to Choosing Data Collection Methods References and Related Reading

Related Posts

Continuous Data Collection

Continuous data collection methods measure every occurrence of a behavior. These methods either count each instance of the behavior or the specific amount of time a behavior occurs. These methods allow you to measure behavior along the basic dimensions and accurately detect change in the behavior. Choose a continuous method when programming requires a complete record of behavior. Several different systems provide continuous data.

The video below provides a brief overview of 4 types of continuous data collection methods.

Frequency

Frequency provides a simple count of the behavior that occurs. Record the frequency of the behavior using tally marks, a clicker, or even small objects. Moving small beads from one pocket to another when a behavior occurs provides a simple way of discretely counting the occurrence of the behavior. Once the session ends, count the number of beads and enter that number onto your data sheet. Use frequency measurement when the behaviors:

  • Have a clear beginning and end
  • Occur at a rate that can accurately be counted

Frequency measures both desirable and undesirable behavior such as the number of times your client pees on the potty or throws toys. Simple frequency counts are easy for staff to collect; however, they don’t take into consideration how long each session is and can be misleading when session duration varies widely. For example, 10 instances of behavior occurring in a 30-minute session is quite different than 10 instances occurring during a 4-hour session.

Record frequency data using tally marks or use a clicker to count then record the total on a simple frequency data sheet. Download the simple frequency data sheet below to get started.

ABA Frequency Data Sheet

Rate

Rate measurements level the playing field regarding session duration. They reflect the frequency of behavior that occurs over a period of time. This type of data provides you with more information than frequency data alone and typically is a more accurate representation of the behavior. To calculate rate, divide the frequency by the duration of the session (i.e. minutes or hours). Rate is expressed as a number per unit of time (i.e. 6 instances per hour or 12 instances per minute).

Use rate data when:

  • Session duration is inconsistent
  • You measure the behavior during some parts of the day but not others

As with other continuous data collection measures, rate data can be used to document behaviors targeted for increase or decrease. You may choose to use rate data to determine the number of times per hour your client mands for something he wants or engages in aggressive behaviors.

Record rate data in much the same way that you record frequency data, but specify the period of time the behavior occurred during (i.e. per minute, per hour, per day, etc.). Download the simple rate data sheet below to get started.

ABA Rate Data Sheet

Duration

Duration data measure how long a behavior lasts from beginning to end. When defining behaviors to be collected by duration recording, professionals must identify an onset and offset to ensure accurate measurement. A stopwatch or timer in a data collection app provides the most reliable duration data. Avoid estimating duration by expecting staff to look at a clock or watch to measure the time. Use duration recording when the behaviors:

  • Have ambiguous beginning and end
  • Last for an inconsistent period of time

Again, duration could measure both behaviors you want to increase as well as those you are looking to decrease. Amount of time spent engaging in imaginative play or in a tantrum are examples of behaviors you might measure using duration data.

Record duration data by noting the amount of time a behavior occurred during each occurrence. You can also calculate the frequency of the behavior by counting the number of times you recorded the duration. Download the simple duration data sheet below to get started.

ABA Duration Data Sheet

Latency

Latency measures the time between the discriminative stimuli (SD) and the response. This measure allows you to evaluate the speed of responding to a particular stimulus. For example, you may use latency data to increase the rate of responding during DTT (by decreasing latency) or you may use it to decrease the rate of responding prior to hearing the full SD during DTT (by increasing latency). Use latency data when:

  • Responses occur too slowly or too quickly following the SD

Latency measures provide highly specific information. To record latency, initiate the time on a stopwatch following the delivery of the SD and stop the time as soon as the learner begins to respond.

Record latency data by recording the SD and the amount of time it took the learner to begin the response. Download the simple latency data sheet below to get started.

ABA Latency Data Sheet

Interresponse Time

Interresponse time (IRT) measures the time between consecutive responses or the end of one instance of the behavior and the beginning of the next. This measurement allows you to measure how quickly the learner engages in multiple instances of the same behavior. For example, you may use interresponse time to increase how quickly a learner completes a math worksheet with 20 problems on it (by decreasing the interresponse time between math problems) or decrease how quickly a learner eats a meal (by increasing the interresponse time between bites of food).

ABA IRT Data Sheet

Choosing Between Types of Continuous Data Collection Methods

When deciding on a continuous data collection method to use to measure a specific behavior, you need to consider several factors to ensure you choose the most appropriate method for your specific situation. Here are some considerations:

  1. Behavior being measured: The nature of the behavior you are tracking will influence your data collection method. Different behaviors may require different measurement techniques. For example, frequency recording may be suitable for tracking instances of a specific behavior, while duration recording may be more appropriate for behaviors that occur over a period of time.
  2. Measurement goals: Clarify your measurement goals and what specific information you need to gather. Are you interested in the frequency, duration, intensity, latency, or some other aspect of behavior? Different measurement methods are designed to capture different dimensions of behavior, so align your goals with the appropriate method.
  3. Feasibility and practicality: Consider the practicality and feasibility of different data collection methods given your available resources, staff experience, and the environment in which data collection will take place. Some methods may be more time-consuming or require specialized understanding, whereas others may be more straightforward and less resource-intensive.
  4. Reliability and accuracy: Think about the reliability and accuracy of the data collection method. Consider the potential for observer bias, human error, limitations of online data collection tools, or interference with the behavior being measured. Choose a method that minimizes these potential sources of error and maximizes the reliability of the data collected.
  5. Individual and setting factors: Take into account the characteristics of the learner or group you are working with, as well as the specific environment or setting in which the behavior occurs. Certain data collection methods may be better suited to different learners or settings.
  6. Ethical considerations: Ensure that the chosen data collection method respects ethical considerations and maintains the privacy and dignity of the individuals involved. Choose methods that minimize intrusiveness and are respectful of privacy, culture and confidentiality.

With the above considerations in mind, let’s look at the specific uses for each of the continuous data collection methods. The image below depicts a comparison between the different types of continuous data collection methods. The description below will help you interpret the image.

Comparison between  types of continuous data collection methods in ABA
Comparison between types of continuous data collection methods in ABA
  • Latency is represented as the time between the SD and the first response. In this example, 4 seconds.
  • Duration is depicted as how long each response lasted. In this example, 9 seconds, 7 seconds, and 10 seconds. This is a total duration of 26 seconds or an average duration per occurrence of 8.6 seconds.
  • The frequency is the number of times the response occurred. In this example, the frequency is 3.
  • The rate is the number of time the response occurred over the period of time. In this example, the rate is 3 per minute.
  • The interresponse time is represented by the time between responses. In this example, 3 seconds and 6 seconds which is an average IRT of 4.5 seconds.

Scroll through the images below to see each individual method separately.

There is not one “right” data collection method for a specific behavior. Choose a method that provides you with the information you need to determine if the intervention is sufficiently effective or if you need to change your strategies. Here are some general guidelines for each of the data collection methods:

  • Choose latency if you’re primarily interested in how long it takes the learner to respond. This may include:
    • Time from antecedent (trigger) until learner engages in target behavior
    • Time from teaching instruction until the learner initiates following the instruction
    • Time from sitting at the table until the learner initiates eating
  • Choose duration if you’re primarily interested in how long a behavior lasts. This may include:
    • Crying
    • Tantrums
    • Aggressive episodes that last for extended periods
    • Time spent reading
  • Choose frequency if you’re primarily interested in how often a behavior occurs and the observation times remain a constant length of time. This may include:
    • Number of times flopping occurs during a school day
    • Number of mands that occur in each 3 hour session
    • Number of times a learner initiates an interaction with a peer during a 20 minute recess period
  • Choose rate if you’re primarily interested in how often a behavior occurs and the observation times vary in length. This may include:
    • Number of times the learner throws an object not intended to be thrown during a session
    • Number of times the learner had toileting accidents between school and the time he went to bed
    • Number of times the learner called out in class when the school has rotating double block periods so some class times are longer than others
  • Choose interresponse time if you’re primarily interested in how much time passes between consecutive responses of the same behavior. This may include:
    • Time between consecutive bites of food
    • Time between getting out of the chair during class
    • Time between incidents of SIB

The above are simply examples of when you might choose to use each of the measurement systems. The measurement system you choose might vary if, for example, when the learner flops he tends to remain on the ground for extended periods that vary in length. If this were the case, duration data might be a more appropriate measurement.

Want a resources that will help you conduct an FBA and create a function-based BIP? Check out our Master ABA Dojo membership! *

Back to Top

Discontinuous Data Collection

Discontinuous data collection systems measure only a sample of behavior that occurs by breaking the session down into small increments of time. Although these data are potentially less accurate than continuous data collection methods, they are easier to collect in busy environments. There is an inherent error in each method of discontinuous measurement (Fiske & Delmolino, 2012). When choosing a discontinuous measurement method, consider these errors carefully.

The chart below describes these errors.

Discontinuous Measurement MethodType of ErrorUse for:
Partial Interval RecordingOverestimates the occurrence of behaviorBehavior decrease
Whole Interval RecordingUnderestimates the occurrence of behaviorBehavior increase
Momentary Time SamplingNeither overestimates nor underestimates the occurrence of behaviorHigh frequency, behavior increase
Comparison of the errors inherent in discontinuous measurement methods

Watch the video below for an example of the errors created by each of these methods.

Partial Interval

Partial interval data breaks the session into equal parts (intervals). Record if the behavior occurred at any point during that interval. Since the behavior only needs to occur once or for a small fraction of the interval, partial interval data overestimates the occurrence of behavior. Use partial interval recording when:

  • The behavior does not have a clear start and stop
  • The behavior occurs at such a high rate that it’s impractical to attempt to count each occurrence
  • An estimate of the frequency of the behavior is acceptable

Keep in mind that because partial interval data provides an overestimate of the occurrence of behavior, you want to use the smallest interval that is practical for your situation. The larger the interval, the more inflated the data. Often, due to the overestimation of this method, professionals use partial interval to document behaviors targeted for reduction. Examples of behaviors you might record using partial interval data include the occurrence of stereotypies or screaming across an entire day, if either behavior occurs at a high rate.

Because staff only need to attend to the behavior if it occurs, partial interval recording can be more efficient for busy staff to collect than monitoring for and counting each occurrence of a given behavior.

Whole Interval

Whole interval data again breaks the session into equal parts (intervals). Record if the behavior occurs throughout the whole interval. Since the behavior must occur for the entire amount of the interval, this method underestimates the occurrence of the target behavior. Use whole interval recording when:

  • The behavior occurs over long periods of time
  • It’s impractical to use duration recording in your setting
  • An overestimate of the behavior is acceptable

Because this method underestimates the occurrence of the behavior you want to use the smallest interval that is practical for your situation to ensure the most accurate reflection of the behavior. Often, due to the underestimation of the occurrence of the behavior, professionals use whole interval to document behaviors target for increase. This might include behaviors such as amount of time spent engaged in table work or functional play, assuming that these occur over significant periods of the child’s day.

Momentary Time-Sampling

Momentary time-sampling takes a quick snapshot of whether or not a behavior occurs. Identify an appropriate interval based on baseline data. When the interval is over, record whether or not the behavior is occurring at that time. This data collection method neither over nor underestimates the behavior; however, because not every instance of the behavior is recorded, the data are far less accurate than continuous data collection. Use momentary time-sampling when:

  • Other methods of data collection are impractical in your situation
  • You rely on someone else to collect the data who is unable to continuously monitor the behavior due to other responsibilities
  • It’s not necessary to ensure you get a complete recording of the behavior

Momentary time-sampling does not provide you with reliable data; however, busy professionals are able to collect data about behavior when they don’t have the ability to attend to the child for extended periods of time. Momentary time-sampling may provide sufficient information for behaviors such as working independently at school or playing alone at home.

Interval data sheet

Choosing Discontinuous Data Collection Methods

Selecting an appropriate discontinuous data collection method depends on the specific clinical objectives, the behavior being measured, and the practical constraints of the situation. Discontinuous data collection methods are typically used when continuous measurement is not feasible. Here are some considerations to help you decide which discontinuous data collection method to use:

  1. Partial Interval Recording: This method involves dividing the observation period into short intervals and recording whether the behavior occurred at any time during each interval. It is suitable when you are interested in estimating the occurrence of behavior within a given time frame. Partial interval recording tends to overestimate behavior occurrence so it is best used for behaviors targeted for reduction.
  2. Whole Interval Recording: In this method, you record whether the behavior occurred continuously throughout the entire interval. It is appropriate when you want to ensure a higher level of accuracy by documenting if the behavior is maintained throughout the entire interval, but requires someone to watch the learner for the entire interval. Whole interval recording tends to underestimate behavior occurrence so it is best used for behaviors targeted for increase.
  3. Momentary Time Sampling: This method involves recording whether the behavior is occurring at the exact moment when each predetermined interval ends. It is useful when the person responsible for recording data has many other responsibilities.

When selecting a method, consider the following factors:

  • Behavior characteristics: Different data collection methods are more suitable for different behaviors. Consider whether the behavior is continuous or discrete, the duration of the behavior, and whether it occurs in bursts or intermittently.
  • Resources and practical constraints: Consider the available time, personnel, and other resources. Some methods may require more intensive data collection efforts than others.
  • Research or clinical goals: Clarify your objectives. Are you interested in overall occurrence, patterns, or specific points in time when the behavior occurs?

Remember to choose a method that is reliable, practical, and aligns with your specific objectives.

Back to Top

Other Data Collection Methods

The above data collection methods provide the widest application for learning about the occurrence of an identified behavior. The list above meets most of the data collection needs of professionals in the field of ABA. Other methods capture information missed when using those methods. The below data collection systems are used less frequently and for more specific purposes than the methods listed above.

ABC Data Collection

ABC data is often a critical component when conducting a functional behavior assessment (FBA). This data collection method looks at what happens right before and right after the behavior you’re interested in. This allows for analysis of the context of the behavior to begin to determine a possible function. Although there are many ways to collect ABC data, one simple method is to create a form with checkboxes for commonly occurring antecedents, behaviors, and consequences. Not only does this make data collection simpler, but it also provides an easier method for analyzing these data.

ABC data sheet

For more information on ABC data, see our posts: ABC Data: The Key to Understanding Behavior and Functions of Behavior in ABA: Complete Guide.

Scatterplot

A scatterplot provides information of the occurrence of behavior across different parts of the day, either time frames or activities. This method allows for a visual analysis to determine if patterns exist. The example below breaks the day into 1 hour blocks of time and then provides space to compare data across an entire week. This data sheet also provides space to document location. This allows you, at a glance, to see that the highest rate of behavior occurs from 7-8 pm at home. While it doesn’t give you a specific count of behavior, this information allows you to determine when you should look to collect that more specific data.

Scatterplot data sheet example

Permanent Product

Permanent product data provides a way for the professional to evaluate the occurrence of behavior after it has stopped. With this method of data collection, the professional does not need to be available to observe the behavior as it occurs. Schools use a lot of permanent product data recording for this reason. A teacher is unable to observe each of her students as they work, but she can look at the permanent products they produce (i.e. worksheets, projects, videos, etc.).

Probe

Probe data simply test to determine if a behavior occurs or does not occur in a given situation. In an effort to ensure that data collection does not interfere with teaching methods, a professional may choose to utilize a probe only data collection system for specific targets. If the professional chooses to do a probe prior to any teaching trials, we refer to this as a “cold probe.” Probe data allow for the professional to focus her attention on teaching methodology including errorless learning and prompt fading techniques. This data system is most beneficial when there are a limited number of individuals working with a client and who don’t rely on the data to know the correct prompt level to use during teaching.

Back to Top

Advantages and Disadvantages of Each Type of Data Collection Method

Each data collection method has its own unique advantages and disadvantages. Consider these carefully before determining which system to use.

AdvantagesDisadvantages
Continuous Data
Collection
~Most accurate
~Sensitive to small
changes in behavior
~Requires constant
observation
~Difficult to use without
1:1 staff
Discontinuous Data
Collection
~Easier to use in a
busy environment
~Provides enough
information for
many situations
~Able to track very high
frequency behavior
~Only an estimate of
behavior
~Must consider over or
under estimation when
analyzing data
~May need more time to
see changes in behavior
Comparison of the advantages and disadvantages of continuous and discontinuous data collection methods
Back to Top

5 Ways to Make Frequency Data Collection Easier and More Discrete

Identifying the obstacles to accurate data collection helps reveal creative solutions to over come them. While you may experience specific obstacles in your practice, two obstacles seem to occur across various types of ABA programs. Often staff have the most difficulty in accurate data collection when they need to collect frequency data for high frequency behaviors or when the act of collecting the data becomes reinforcing for the child.

High Frequency Behaviors

Many children with autism engage in a variety of maladaptive behaviors that occur at too high a rate. Often behavior change occurs gradually over time. In order to determine the effectiveness of a behavior reduction plan, the data that are collected must be accurate. Although frequency and rate data collection may not be the most practical method for collecting data for high frequency behaviors, it is the most sensitive to behavior change, provided that the data are accurate.

Data Collection Becomes Reinforcing

Many children with autism don’t pay attention to data being collected by a professional. Other children find any form of attention motivating. When these children associate their behavior with the behavior of staff, the result may be an increase in this behavior. Traditional data collection techniques require staff to distinctly make some type of mark or electronic record. These methods lack the subtlety needed for children attuned to staff behavior.

Data Collection Techniques

Here are 5 options for easier and more discrete data collection. Each of these techniques offers its own unique advantages and disadvantages. Try them and see what works best for you!

1. Clicker Counters

Clicker counters are a great tool for counting high frequency behaviors and can easily be used to calculate rate (just calculate frequency/time). These clickers are a great way to track a variety of behaviors that occur frequently. You simply assign each behavior a color and click when the behavior occurs. They include a hook that you attach to a carabiner so you can attach them to a pocket or belt loop. These are affordable and available on Amazon . Their major disadvantage is the clicking sound that they make. Avoid these for children whose behavior is reinforced by staff collecting data.

Alternatively, digital finger counters provide the subtly you may need, but these counters can be sensitive and you risk counting behavior that doesn’t occur. If worn on a thumb, you may be less likely to accidentally hit the button. Again, these handy tools are inexpensive and easy to find on Amazon .

2. Small Objects in Pockets

Check out this low-tech option for tracking the occurrence of behavior. Put a collection of small objects in one pocket. As behavior occurs, transfer the corresponding number from that pocket into the other pocket.

This method offers some subtlety for well trained staff. To optimize discretion, staff put their hand in their pocket and carefully scoop one item into their hand as behavior occurs. Once the child looks away, staff move the items from that pocket to the other. The more frequently the behavior occurs, the smaller the objects must be.

With this option, you risk accidentally dropping the items either back into the original pocket or on the floor when transferring them. In addition, you must count each item at the end of the session. While this creates additional work, it may be the best short-term option if you need a low-tech, discrete data collection method.

3. Beads on a Pipe Cleaner

Sliding beads on a pipe cleaner offers another low-tech option for frequency data collection. When done well, it appears as though staff are simply fidgeting with the beads or doing a craft while not attending to the child’s behavior. Staff can slide a group of beads onto the top of the pipe cleaner and as behavior occurs, slide the beads to the bottom.

While this option may be reinforcing for staff as it can offer a calming effect for some, children may pick up on the fact that each time behavior occurs, staff pick up the beads. There are 2 ways to combat this effect: rotate between different low-tech options or have staff play with the beads throughout the day when behavior does not occur.

4. Technology

Technology offers a broad array of options from simple to complex. Many apps provide access to different methods to tally behavior as it occurs. Some apps export or graph this data for you as well. A simpler option is to open a note taking app and add an emoji or other character each time the behavior occurs.

Similar to the small objects method, this method requires staff to count the occurrences at the end of the session. This method offers some other nice advantages for children who attend to staff behavior. Adults on technology (phones, tablets or computers) is so common place that it essentially becomes unnoticeable to many children. In addition, children who are motivated by technology may be unpleased to see that the adult has found her own entertainment while he engages in the behavior.

5. Small Elastics on Fingers

A final low-tech and inexpensive option is to place small elastics on your fingers (hair elastics intended for young children work well). Place spares on one hand and as behavior occurs, roll them over to the other hand. To the child, this may appear as simple fidgeting.

This method may be inefficient for behaviors that occur at a very high frequency. In addition, some staff may find them uncomfortable on their fingers. This system also requires staff to count each elastic at the end of the session. Despite these disadvantages, you may find that this method works best for your specific circumstances.

Back to Top

Research in Data Collection Methods

Several studies evaluate the use of the different data collection methods in research studies. Two studies evaluate this trend during different time periods. Kelly (1977) looked at the research published in The Journal of Applied Behavior Analysis from 1968-1975 and Mudford, Taylor, and Martin (2009). The results of their research are presented in the table below.

Time PeriodContinuous Data Collection MethodsDiscontinuous Data Collection Methods
1968-197559%41%
1995-200555%45%
Comparison of the use of continuous versus discontinuous data collection methods in the research

Across the 40 year span from the start of the first study to the end of the follow-up study, researchers balanced the use of continuous and discontinuous data collection methods. Although discontinuous data collection methods have their value, they fail to quantify the basic dimensions of behavior (Fiske & Delmolino, 2012). With the introduction of electronic data collection systems, continuous data measures are more practical and efficient than they were when only paper and pencil technology existed. Although this should lead to researchers relying more on continuous data collection measures, the studies reveal that it’s not necessarily the case.

Factors to Consider When Choosing Between Continuous and Discontinuous Data Collection Methods

When determining which data collection method fits your needs, you must consider many factors including:

  • Is the behavior potentially dangerous?
  • Does the behavior threaten the placement of the individual?
  • How frequently does the behavior typically occur?
  • Does the behavior occur over a period of time?
  • Does the behavior have a clear beginning and end?
  • Who will be collecting the data?
  • How long does the behavior usually last?
  • Is it a behavioral deficit or excess?

Fiske and Delmolino (2012) provided clear guidelines for choosing between a continuous and discontinuous data collection method. The table below is a description of their recommendations.

Continuous Data Collection MethodDiscontinuous Data Collection Method
Discrete behaviors with a clear onset and offsetAmbiguous breaks between the occurrence of the behavior
Interventionist can accurately record each instance The behavior occurs at a very high rate
Recording behaviors individuallyRecording multiple behaviors simultaneously
Interventionist responsible for 1 learnerInterventionist required to complete many tasks at once
When to choose a continuous or discontinuous data collection method

Serious, dangerous or severe behavior requires a system that provides accurate data. Strongly consider using continuous data collection when addressing these types of behavior. Behaviors that occur at an exceptionally high rate may require a discontinuous data collection method for accuracy. If you rely on parents or teachers to collect data, you should consider discontinuous methods. Taking all of these factors into consideration allows you to select the most effective and efficient data collection system for your ABA program.

Back to Top

Ethical Considerations When Choosing Data Collection Methods

The table below presents some important ethical considerations when choosing data collection methods. The table includes specific action steps to help you ensure you practice in an ethical way.

Ethical ConcernDescriptionAction Steps to Ensure Ethical Practice
FeasibilityData collection systems must be feasible to use in the setting where they will be used.– Implement strict protocols for data storage, access, and sharing to protect participants’ privacy. <br>- Anonymize or de-identify data whenever possible to reduce the risk of identifying individual participants.
Validity and ReliabilityEnsuring that data collection methods accurately measure what they intend to measure and produce consistent results over time and across different observers.Consider the resources available in the setting when making a decision about which data collection system to use.
Cultural SensitivityConsidering cultural norms, values, and practices when selecting data collection methods to avoid potential harm or discomfort to participants from different cultural backgrounds.– Conduct a cultural assessment to understand the cultural context and preferences of participants.
– Modify data collection methods to be culturally sensitive and respectful.
Bias and ObjectivityMinimizing bias in data collection by maintaining objectivity and avoiding personal or professional biases that may influence the collection or interpretation of data.– Train data collectors on the importance of objectivity and the potential for bias.
– Implement inter-rater reliability checks to ensure consistency and minimize bias between different observers.
Transparency and Full DisclosureProviding transparent and complete information to participants about the purpose, scope, and potential uses of the collected data, including any data-sharing practices or data ownership rights.– Clearly communicate to participants how their data will be collected, used, and shared.
– Obtain explicit consent for any secondary use or sharing of data beyond the original purpose.
Voluntary ParticipationEnsuring that participation in data collection is voluntary and participants have the right to withdraw their participation at any time without negative consequences.– Clearly communicate to participants that their participation is voluntary and they can withdraw at any time without penalty.
– Respect participants’ decisions to withdraw and provide a straightforward process for withdrawal.
Minimizing HarmTaking measures to minimize any potential harm or negative consequences to participants during the data collection process.– Conduct a risk assessment to identify potential sources of harm and implement measures to mitigate those risks.
– Monitor participants closely for any signs of distress or discomfort and provide support or referrals as needed.
Beneficence and UtilityEnsuring that the data collection methods and resulting data provide meaningful insights, benefits, or contributions to the field of study or the participants themselves.– Conduct a thorough review of the potential benefits and utility of the data collection methods.
– Regularly evaluate the impact and usefulness of the data collection methods and make adjustments if necessary.
Data Integrity and AccuracyEnsuring that data is collected accurately, recorded appropriately, and protected from any intentional or unintentional alterations or falsifications.– Implement data quality checks and validation procedures to verify the accuracy and integrity of collected data.
– Protect data from unauthorized access or tampering through secure data storage and access controls.
Transparency in Data AnalysisConducting data analysis with transparency, avoiding selective reporting or manipulation of data to fit predetermined hypotheses, and ensuring a comprehensive and unbiased interpretation of the results.– Follow established data analysis procedures and guidelines to ensure transparency and rigor.
– Clearly report the methods, results, and limitations of the data analysis to promote transparency and reproducibility.

By considering and implementing these action steps, BCBAs can promote ethical practice in choosing data collection methods and ensure the protection, rights, and well-being of the learners they serve.

Back to Top

Below is a table summarizing research articles related to choosing data collection methods. The table includes important action steps to help you put these ideas into practice.

Article TitleSummaryAction Steps for Applying Information
Use of discontinuous methods of data collection in behavioral intervention: Guidelines for practitionersThe article provides guidelines for practitioners on using discontinuous methods of data collection in behavioral interventions. It discusses considerations for choosing appropriate data collection methods, including partial interval recording, whole interval recording, and momentary time sampling.Consider the context and objectives of the intervention when selecting a data collection method. Ensure practitioners are trained in using discontinuous methods accurately and consistently. Regularly review and update data collection procedures based on the specific needs of the individuals receiving the intervention.
An evaluation of estimation data collection when measuring problem behavior in a classroom settingThis article evaluates the use of estimation data collection for measuring problem behavior in a classroom setting. It compares the accuracy and efficiency of estimation data collection with direct observation and discusses the implications for practitioners.Consider the feasibility and practicality of using estimation data collection in classroom settings. Assess the accuracy and reliability of estimation data collection by comparing it with direct observation methods. Determine if estimation data collection can provide sufficient information for decision-making and intervention planning.
A review of the observational data‐collection and reliability procedures reported in the Journal of Applied Behavior AnalysisThis article reviews observational data collection and reliability procedures reported in the Journal of Applied Behavior Analysis. It provides insights into common practices and recommendations for improving observational data collection and reliability.Familiarize yourself with established observational data collection procedures. Implement systematic training for observers to enhance reliability. Regularly assess interobserver agreement and make adjustments as needed. Consider using validated data collection tools and software to improve accuracy and efficiency.
Data collection and measurement assessment in behavioral research: 1958–2013The article presents an analysis of data collection and measurement assessment practices in behavioral research from 1958 to 2013. It highlights trends, challenges, and recommendations for improving data collection and measurement procedures.Stay updated on evolving data collection and measurement assessment practices in the field. Implement evidence-based techniques for data collection and measurement. Regularly review and adapt data collection procedures to align with current best practices. Seek professional development opportunities to enhance knowledge and skills in data collection and measurement assessment.
Procedures and accuracy of discontinuous measurement of problem behavior in common practice of applied behavior analysisThis article examines the procedures and accuracy of discontinuous measurement of problem behavior in applied behavior analysis. It provides recommendations for practitioners to improve the reliability and validity of discontinuous measurement methods.Ensure practitioners are trained in accurate implementation of discontinuous measurement methods. Regularly assess and monitor the reliability and validity of discontinuous measurement data. Consider the limitations and potential biases associated with discontinuous measurement and make appropriate adjustments.
The representativeness of observational samples of different durationsThe article investigates the representativeness of observational samples of different durations. It discusses the implications of sample duration on the accuracy and generalizability of observational data.Carefully consider the duration of observational samples to ensure they adequately represent the behavior of interest. Balance the need for detailed data with practical constraints. Conduct pilot studies to determine optimal sample durations based on the specific behavior and context.
Continuous recording and interobserver agreement algorithms reported in the Journal of Applied Behavior Analysis (1995–2005).This article examines continuous recording and interobserver agreement algorithms reported in the Journal of Applied Behavior Analysis from 1995 to 2005. It provides insights into the use of continuous recording methods and their impact on interobserver agreement.Familiarize yourself with various continuous recording methods and algorithms used in behavior analysis. Train observers on accurate implementation of continuous recording procedures. Regularly assess interobserver agreement to ensure consistency and reliability. Adjust continuous recording procedures based on feedback and observations to optimize interobserver agreement.
A comparison of frequency, interval, and time‐sampling methods of data collectionThis article compares frequency, interval, and time-sampling methods of data collection. It examines the advantages, disadvantages, and considerations for each method in different contexts.Evaluate the specific goals and requirements of the data collection task to determine the most appropriate method (frequency, interval, or time-sampling). Consider the behavior being measured and the resources available for data collection. Choose a method that balances accuracy and practicality. Regularly review and analyze the collected data to ensure it aligns with the research or intervention objectives.
A comparison of data collection techniques used with discrete trial teachingThe article compares different data collection techniques used with discrete trial teaching in the context of autism spectrum disorders. It discusses the advantages and limitations of various methods and provides recommendations for practitioners.Familiarize yourself with different data collection techniques applicable to discrete trial teaching. Select a method that aligns with the specific goals and characteristics of the individuals with autism spectrum disorders. Ensure accurate implementation of the chosen method through training and ongoing supervision. Regularly evaluate the reliability and validity of the collected data to inform decision-making and intervention adjustments.
Back to Top

References and Related Reading

Fiske, K., & Delmolino, L. (2012). Use of discontinuous methods of data collection in behavioral intervention: Guidelines for practitionersBehavior Analysis in Practice5(2), 77-81.

Griffin, A. (2020). An evaluation of estimation data collection when measuring problem behavior in a classroom setting. Behavior Analysis in Practice, 13(4), 724-732.

Kelly, M. B. (1977). A review of the observational data‐collection and reliability procedures reported in the Journal of Applied Behavior AnalysisJournal of Applied Behavior Analysis10(1), 97-101.

Kostewicz, D. E., King, S. A., Datchuk, S. M., Brennan, K. M., & Casey, S. D. (2016). Data collection and measurement assessment in behavioral research: 1958–2013. Behavior Analysis: Research and Practice, 16(1), 19

LeBlanc, L. A., Lund, C., Kooken, C., Lund, J. B., & Fisher, W. W. (2020). Procedures and accuracy of discontinuous measurement of problem behavior in common practice of applied behavior analysis. Behavior Analysis in Practice13(2), 411-420.

Mudford, O. C., Beale, I. L., & Singh, N. N. (1990). The representativeness of observational samples of different durationsJournal of Applied Behavior Analysis23(3), 323-331.

Mudford, O. C., Taylor, S. A., & Martin, N. T. (2009). Continuous recording and interobserver agreement algorithms reported in the Journal of Applied Behavior Analysis (1995–2005)Journal of Applied Behavior Analysis42(1), 165-169.

Repp, A. C., Roberts, D. M., Slack, D. J., Repp, C. F., & Berkler, M. S. (1976). A comparison of frequency, interval, and time‐sampling methods of data collectionJournal of Applied Behavior Analysis9(4), 501-508.

Taubman, Mitchell T., Ronald B. Leaf, John J. McEachin, Sasha Papovich, and Justin B. Leaf. “A comparison of data collection techniques used with discrete trial teaching.” Research in Autism Spectrum Disorders 7, no. 9 (2013): 1026-1034.

Whoa, hold on there!

This course is NOT available outside our membership

…unless you subscribe.

"*" indicates required fields

First Name*
This field is for validation purposes and should be left unchanged.

THE BALANCED BCBA

The key to breaking free from billable hours

and insurance companies.

"*" indicates required fields

First Name*
This field is for validation purposes and should be left unchanged.

Wait! You almost missed it!

This course is NOT available outside our membership

…unless you subscribe.

"*" indicates required fields

First Name*
This field is for validation purposes and should be left unchanged.