Monitoring Progress of Intensive Behavior Supports

Intensive behavior supports require close and thoughtful monitoring. Students who engage in extensive interfering behaviors can take time to respond to a BIP or other support practices. Daily progress monitoring in the first stage of a BIP’s implementation can help teams distinguish between a BIP that is ineffective versus a BIP that is not yet fully effective.

When a BIP appears ineffective after 4–6 weeks of implementation, teams should consider two critical questions:

Is the BIP fully implemented?
Is the BIP a good fit for the student?

To consider these critical questions, teams must have gathered and graphed behavior progress data to review together. Strategies for data collection are described below.

Choosing the Right Tool for Data Based Decision Making

There are a variety of ways to measure a student’s behavior. To select a measurement tool, teams should consider the characteristics of the behavior they are measuring, and the question they are seeking to answer. Identifying the best measurement method depends on 1) how frequently the behavior occurs, and 2) whether the behavior is discrete or non-discrete.

Discrete behaviors have a clearly recognizable (and usually rapid) start and stop. Examples include raising one’s hand, kicking a wall, or screaming
Non-discrete behaviors typically have a less clear start and stop, and are often more complex categories of behavior the student engages in for longer periods of time. An example includes off-task behavior – which might include varying degrees of more than one behavior considered off task (e.g., being out of seat, talking to neighbors, looking off into space, sleeping during instruction, working on things not related to the task)

Whatever method of data collection the team uses, that method should be the same across baseline and intervention phases to allow for accurate comparisons between phases.

The sections below summarize a variety of common behavior data collection methods, and includes guidance for their selection.

Frequency Data

Frequency data can answer the question, “How often does this behavior occur?” and are collected by recording a tally mark each time the behavior occurs during data collection. This method is appropriate for discrete behaviors with a clear start and stop, and for behaviors with low to moderate frequency. Behaviors that occur with extremely high frequency (e.g., many times per minute) are usually not practical to record with frequency data.

Duration Data

Duration data can answer the question, “How long does this behavior last?” Data are collected by writing down the start and end time for each instance of the behavior, and then subtracting the earlier time from the later time to get the duration. Data should indicate whether the time measured is in seconds, minutes, or hours. This measure is appropriate for non-discrete behaviors.

Latency Data

Latency data can answer the question, “How long until a behavior occurs?” It involves similar timing procedures as duration recording, and measures the time between an event (e.g., the teacher asking students to start independent work) and the beginning of a desired behavior (e.g., the student beginning their work). This requires careful observation during specific moments. This measure is appropriate when supporting the student to learn certain alternative or replacement skills. It can also help the team measure changes in a student’s prompt dependency (which can develop if the student has been over supported by adults).

Interval Recording: Whole Interval

Whole interval data collection involves scoring the interval with a (+) if the behavior of interest occurred for the entire interval. For example:

A teacher is measuring screaming behavior using intervals that are 15 seconds long. The student screamed for the entire 15 second interval, so the teacher marks that interval with a (+)
A school social worker is measuring on-task behavior using intervals that are 30 seconds long. The student was on task for the entire 30 second interval, so the school social worker marks that interval with a (+)
A paraeducator is measuring out of seat behavior using intervals that are 60 seconds long The student was out of her seat for the first 55 seconds and seated for the last 5 seconds of the interval, so the paraeducator marks that interval with a (-)

This method of interval recording can be appropriate when measuring non-discrete behaviors with a long duration. Whole interval recording is not usually appropriate for behaviors best measured with frequency data (e.g., hitting, swearing). In addition, whole interval recording tends to underestimate behavior since the interval is only scored if the behavior occurs for the entire time. To address this, select a shorter interval length, or consider a different measure (such as duration recording or momentary time sampling).

Interval Recording: Partial Interval

Partial interval data collection involves scoring the interval with a (+) if the behavior of interest occurred at any time during the interval. For example:

A teacher is measuring screaming behavior using intervals that are 15 seconds long. The student screamed once during the interval, so the teacher marks it with a (+)
A school social worker is measuring on-task behavior using intervals that are 30 seconds long. The student is briefly on task for one second at the start of the interval, so the school social worker marks that interval with a (+)
A paraeducator is measuring out of seat behavior using intervals that are 60 seconds long. The student is seated for most of the interval, but stands up once for a couple seconds, so the paraeducator marks that interval with a (+)

This method can be effective with behaviors that are infrequent or hard to anticipate. It tends to overestimate behavior since the interval must be scored if the behavior occurs at all during the interval. To address this, select a shorter interval length, or consider a different measure (such as frequency recording or momentary time sampling).

Interval Recording: Momentary Time Sampling

Momentary time sampling involves scoring the interval with a (+) if the behavior of interest is occurring at the moment the interval ends. For best results, the observer should look away from the student until the last moment of the interval, then look at the student to determine if the behavior is occurring. This allows for the most accurate measurement without allowing the rest of the students’ actions leading up to that moment to influence data collection. For example:

A teacher is measuring screaming behavior using intervals that are 15 seconds long. The student screams twice during the interval, but neither time was at the 15-second mark, so the teacher marks the interval with a (-)
A school social worker is measuring on-task behavior using intervals that are 30 seconds long. When the school social worker glances up at the end of 30 seconds, the student is on task, so she marks that interval with a (+)
A paraeducator is measuring out of seat behavior using intervals that are 60 seconds long. The student is seated when the paraeducator looks up at the 60-second mark, so he marks that interval with a (-)

This method balances some of the disadvantages of whole interval and partial interval recording. It is appropriate for a variety of behaviors, but still carries the risk of underestimating behavior if intervals are inappropriately long.

Behavior Rating Scales

A behavior rating scale is another method of estimating behavior change. Teams can create a simple scale based on their knowledge of the student and their specific behavior patterns. Data should be collected frequently enough that it can inform decision making. For example:

A team wants to collect baseline data for a student’s behavior using a 3-point rating scale. Using this scale, the teacher circles the appropriate rating on a datasheet at the top of each hour. The points on the scale are defined as follows:

Last hour, the student followed classroom expectations with 0–2 reminders
Last hour, the student followed classroom expectations with 3–4 reminders
Last hour, the student received 5+ reminders to follow classroom expectations

Teams may also use the Direct Behavior Rating (DBR)⁹³ scale. Using this scale, teachers can estimate the occurrence of a behavior using a 0–10 scale. A rating of 0 estimates that the behavior occurred 0% of the time, while a rating of 5 estimates the behavior occurred 50% of the time, and so forth. Since teachers may disagree on their subjective rating using DBR, it works best for showing changes in an individual teacher’s ratings over time.

While less precise than other measures, a rating scale is very feasible for teachers and other staff. It can be a good fit when measuring a student’s use of desired behaviors, or when a student’s specific expression of the interfering behavior changes frequently. It’s important to predefine the points on the rating scale using definitions that are as measurable and observable as possible. All team members should have a shared understanding of which end of the scale corresponds to desired behavior. Teams should also decide in advance when and how often a rating should be recorded.

⁹³ University of Connecticut (n.d.). Direct Behavior Ratings.