Page Links:

Data Recording

There are two primary ways to quantify behavior. The first is to measure how often a behavior occurs (frequency) and the second is to measure how long each behavior lasts (duration).  At times it may also be important to consider measuring other dimensions so behavior such as the time between a request and the start of the behavior (latency) or the power of the behavior (intensity). Staff will need to determine which of these methods is most appropriate for the behavior they are monitoring. In addition to quantifying the behavior, it is often helpful to document the contextual events that typically occur just before (antecedent) or just after (consequence) a behavior. Collecting and using data to guide interventions for behavioral needs does not need to be complicated and should not delay the student in receiving support.

Some data recording methods provide an exact measure of a behavior’s occurrence, while others provide a general estimate or proportion of the behavior’s occurrence. The data recording method selected is best determined by what the behavior looks like (topography), the frequency of the behavior, and the time and effort that staff can dedicate to observation and recording. When more structured and systematic data recording methods are used, the information is more valid if used on a regular basis to make decisions about effectiveness of interventions.

Each data recording method has advantages and disadvantages. It may be most efficient to use a variety of measurements to get the useful information. Some methods are more conducive to the school setting and schedule. The following questions may help guide in the selection process:

What do you want to measure?

You should use:

Does the behavior generate a product (for example, a written assignment, a clean desktop, or completed checklist)?

Permanent Product

Can you easily count every time that the behavior occurs (for example, raising your hand)?

Can you easily identify when the behavior starts and ends?

Event Recording/Frequency

Does the behavior occur so frequently that it may be difficult to count accurately?

Momentary Time Sampling

Interval Recording

Do you want to measure how long it takes for a behavior to begin?



ABC Analysis Sheet

ABC Data Checklist

Do you want to measure how long the behavior lasts?


Do you want to measure progress on multi-step tasks?

Task Analysis

Permanent Product Measurement

Permanent products refer to the real or concrete objects or outcomes that result from a behavior.  Common examples of permanent product measurements are the number of completed items, creative projects, quizzes, homework assignments turned in by student. In other situations, it might be the number of tokens earned or points awarded for following the class rules. Each permanent product should be labeled with the date completed and the number of opportunities the behavior could have occurred and the number of times the behavior did occur.

A benefit of this measurement is that staff does not need to be consistently present to observe the occurrence  and of the behavior. They can also utilize this method without making major changes or additions to their daily routine/schedule. Permanent products can also be stored or filed for later review or verification as needed making this method generally effective and convenient.

A disadvantage of permanent product measurement is that staff may not have a way of knowing the context of how the student engaged (or not) in the behavior. For example, staff may not have the information that few math problems were completed in class today because the student was distracted by a peer sitting nearby or had to wait for requested assistance. Sometimes this method is best used in combination with more direct observation.

Summarizing permanent product measurements can often be calculated as the percentage of products completed given the number of opportunities provided.

Download Permanent Product Recording Sheet

Event Recording/ Frequency

Event recording is a process for documenting the number of times a behavior occurs. Each time a student engages in the target behavior it is counted in some way. This may include marking tally marks on a designated data collection form, using a counter, or using a tangible object to track the number of events (for example, moving a paper clip from one pocket to another). Behaviors measured using event recording should be ones that occur for short periods of time and have a observable beginning and end.

An advantage of using event recording as a measurement is that it can be done unobtrusively and does not need to interfere with instruction. It can be done for a designated short period of time across multiple days to establish a pattern of behavior.  A rate of occurrence (for example, number of behaviors per minute or hour) can be determined that allows you to see a decrease in the occurrence of the target behavior as a result of your intervention.

Disadvantages of this measurement are that it cannot easily be used for high frequency behaviors or with behaviors that occur over extended periods of time (for example, staring out the window).

Summarizing event recording data is done quite simply be graphing the frequency data. The graph will allow staff to monitor the trend toward increasing the desired behavior and/or decreasing the challenging behavior.

Download Event Recording/Frequency Sheet

Download the Daily Event Recording for Multiple Students

Interval Recording/Momentary Time Sample

Momentary time sampling is an example of interval recording. It involves observing whether the behavior occurs or does not occur during specified time periods. That time period is broken into intervals for observation and recording. For example, a 10 minute time period may be broken into 30 second intervals resulting in 20 opportunities to observe for the behavior. The observer looks up at the END of each interval and marks whether the behavior is occurring or not. A timer is necessary to indicate the interval lengths.

A major advantage of this measurement is that you do not need to be observing the student’s behavior all the time. This method can provide you with an estimate of behavior rather than frequency count of the behavior. The shorter time needed for observation make it easier to implement within the daily schedule. It is also an option when it is harder to determine when a behavior begins and ends or when the frequency is too high to count (for example, on task behavior). This method also allows staff to observe more than one behavior or student at time.

A disadvantage is that it can underestimate the rate of occurrence since the behavior may occur during the interval but cease just prior to the moment of observation at the end of the interval.

Summarizing time sampling data is usually presented as percent of intervals of time in which the behavior occurred. To calculate the percent of intervals, divide the total number of intervals in which the behavior was recorded by the total number of intervals during the observation period. For example, Sara was talking to a peer during 12 of the 20 intervals observed; 12/20=60%.

Download Time Sample Recording Sheet

Download Time on Task Recording Sheet

Interval Recording/Scatterplot

The scatterplot measurement is another interval recording method that is typically used to identify the time of the day that the challenging behavior most often occurs. It involves breaking the observation period into a number of smaller time periods or intervals. Staff observe during the observation period and record whether the behavior occurred or did not occur during that interval. This data will reveal any temporal patterns of occurrence (behavior occurs most often between 8:00-8:45 no matter what the schedule) or any pattern that can be associated with specific environmental events (behavior most often occurs during independent math work). This method can be most beneficial in determining the optimal times for someone to directly observe the occurrence of the target behavior and to begin to identify the variables that may be contributing to the behavior.

Advantages of the scatterplot recording is that it takes less time and effort than event recording for tracking high frequency behaviors since you do not need to count each occurrence. The intervals can be set at any reasonable length to allow staff to record data and complete other scheduled responsibilities.

A disadvantage of this method is that it can overestimate the occurrence of the behavior if the intervals are too long.

Summarizing scatterplot data can initially be done by visually scanning the data form for patterns over days/weeks. The typical form is a matrix with the time intervals down the left side of the form and the dates across the top. A key is included that indicates how the observer will mark the each interval (for example, make an X if the behavior occurs or leave blank).  The percentage of intervals in which the behavior occurred is calculated by counting the number of intervals that the behavior occurred divided by the total number of intervals and multiply by 100. Sara was out of her seat 48 out of 130 intervals recorded. 48/130 x 100= 37%.

Download the Interval Recording/Scatterplot Sheet

Duration recording

Duration recording documents the length of a behavior by recording the time from the behavior beginning and ending. This measurement is helpful when your primary concern is the length of time the student engages in the behavior (off task, out of seat, tantrums). It requires an instrument to track time (stopwatch, timer, wall clock) and a form for marking the start and stop times.

An advantage of this measurement is that you sometimes can see gradual improvements in duration before you see changes in frequency of the behavior. This measurement gives you both forms of data. It can be most beneficial when you are measuring a behavior that includes behaviors that are too frequent or harder to measure (pencil tapping, talking to peers).

A disadvantage of this method is that it is sometimes difficult to get accurate measurements unless staff only has the responsibility of observation during this time.

Summarizing duration data includes the calculation of the sums of all the durations divided by the number of occurrences. Sara had three occurrences of out of seat behavior (125 seconds, 365 seconds and 45 seconds) during the observation period. 125 + 375 + 40 = 540/3 =180 seconds or 3 minutes average per out of seat behavior.

Download Duration Recording Sheet

Latency recording

Latency recording is a different form of duration recording that is used when you want to know how long it takes for a student to begin performing a particular behavior once the opportunity has been presented (given a verbal cue or event demand).  For example, if staff request that a student begin working on an assignment, how long does it take for the student to comply with the request. This measurement can be used with behaviors that have a definitive beginning and end. Latency recording requires effort and some way to monitor the time (stopwatch, timer, wall clock).

An advantage of this measurement is that you can see gradual progress in responding more quickly to verbal cues. This can be a beneficial goal for many students. It can also be used to determine exactly when to prompt the use of a new skill or replacement behavior before the challenging behavior occurs.

The disadvantage is obviously the time it takes to monitor while not being able to conduct other classroom obligations.

Summarizing latency data involves calculating the average latency by adding all the data together and dividing by the number of opportunities measured. Sara took 60 seconds, 90 seconds, 50 seconds and 35 seconds to start 4 assignments this week. 60 + 90 + 50 + 35 = 235/4= 58.75 seconds is the average latency for starting on the assignments.

Download Latency Recording Sheet

ABC Analysis

This method of measurement is typically done as part of the Functional Behavioral Assessment. It involves the recording of the environmental variables related to the challenging behavior. When the behavior occurs, the observer records the target behavior, the antecedent (what event preceded the behavior) and the consequence (event that immediately followed the behavior). This assists in identifying the possible function of the behavior. This method is not used for routine daily data collection because it requires the observer to record multiple variables.

Advantage of this measurement during the FBA process is that it provides contextual information about the behavior which can prove helpful in hypothesizing a function or lagging skill contributing to the behavior. Typical information included is date, time, people involved, setting, environmental demands and responses to the occurrence of the behavior.

A disadvantage of this method is that it requires more time and effort by staff and only demonstrates correlation not causation.

Summarizing ABC analysis data results in a frequency count; the total number of times the behavior occurred during the observation period.

Download ABC Analysis Sheet

Download ABC Data Checklist

Task Analysis

This measurement allows staff to track progress made on multi-step tasks. Staff analyze the activity in order to break a complicated skill into smaller skills, e.g. preparing a food item, toothbrushing, rounding numbers. The data allows staff to see small progress being made on more difficult skills and pinpoints where students may need more instruction or support to successfully complete the entire activity. Results will show increase in percentage of steps demonstrated successfully with various amounts of support presented. Results  will also show consistency across time of each step being demonstrated.

An advantage of this measurement is that staff and student note small steps of success which can keep motivation levels higher for continued effort. It also allows staff to examine more complicated skills to provide better instruction and support where data indicates that student may be struggling. It is also an effective way to monitor the level of  independence in performance of the task.

A disadvantage of the task analysis data method is that staff are required to monitor the student engaged in the activity closely to score each individualized step.

Download Task Analysis Recording Sheet

Qualitative Measures

More qualitative or subjective impressions of behavior are generally based on memory and are generally vague and not measurable. It is time consuming to document in this manner and usually does not include all occurrences of challenging behavior. However, these can be used as part of  information review for the Functional Behavioral Assessment process as a supplement to more objective measurements. Incidents involving such interventions as crisis team response or use of physical restraint procedures require written narrative documentation to be done.

Collect qualitative measures using teacher narratives, anecdotal reports, log entries, interviews, surveys and rating scales.

Using data to evaluate the effectiveness of interventions

Data that is collected when monitoring a student’s response to intervention answers the question, “Is this intervention effective?” Without objective measurements, behavior change may be too gradual to determine the student’s response. Just as we would not consider teaching an academic skill without determining the current level of functioning and then monitoring the acquisition of the new skill, we need to apply this same reasoning to behavior skills.

Prior to the start of an intervention with a student, it should be determined what data will be collected, how often will it be reviewed and how it will be visually represented for analysis. The initial data collected establishes the current level or starting point for the student. It is the measuring stick by which intervention data are compared to determine the extent of change that occurred in the behavior. Baseline data also assists in setting challenging yet manageable goals for improvement.

During an intervention, a minimum of eight data points should be collected within three weeks before review. Reviewing progress is most easily done by having the data visually represented. The more data you have collected, the more reliable your decisions about the intervention’s effectiveness. Generally, decisions will include continuing the intervention, intensifying the intervention, modifying the intervention, fading the intervention or returning to the problem solving phase to gather more information.

Graphing your data is important because it allows you to have a visual image of the status of the behavior. A graph allows you to determine at a glance: how often the behavior occurs, when it is increasing or decreasing. Procedures for preparing a graph include labeling the horizontal axis with the time component (days, weeks, sessions) and the vertical axis with the value of the measurement of the behavior (frequency, duration percentage).

Positive response to intervention

If data indicates that the student is making positive progress toward the goal (gap between the trend line and goal line are closing) and will reach the goal within a reasonable amount of time, staff can choose to: 1) continue the intervention with the current goal; 2) continue the intervention with an increased goal; 3) or if progress has been maintained for some time, to begin to teach self-management while fading or withdrawing supports.

Questionable response to intervention

If data indicates that some improvement is seen but the rate of improvement is slow, the response is seen as questionable (the gap between the trend line and the goal line stops widening but does not occur in an acceptable amount of time). The first thing to consider is whether the intervention is being done with fidelity and accuracy (across staff, settings, time). If the intervention has been implemented consistently, then the decision can be made to intensify the intervention or modify the intervention. Take this opportunity to reconsider the function of the behavior, reconsider the goal, provide more frequent feedback, and to allow more student choice and input.

Poor response to intervention

When the data indicates no change (distance between the trend line and the goal line continues to widen), the response to intervention is considered poor. Again, the first consideration is the fidelity and accuracy of the implementation of the intervention. If staff believes the intervention has been provided consistently with no effect, then staff should consider whether the function of the behavior has been correctly identified, if the interventions aligned with this function, or if there are other functions to consider.