Information Brief

Systematic Direct Observation

---

Data Collection Overview

It is often easy for educators to recognize when a student’s behavior is happening and how it is impacting the classroom. However, without data, educators might make incorrect assumptions about how frequently a behavior occurs or how long it lasts based on opinions, feelings, inaccurate recollections, or incomplete information. A systematic process for collecting and analyzing behavioral data in a consistent and accurate manner reduces this subjectivity. Additionally, it helps educators:

• Identify patterns related to when, where, and how behavior occurs
• Make informed decisions about how to address behavior
• Determine if behavioral supports and interventions have been effective
• Decide if an intervention should be continued, adapted, or stopped
• Monitor student progress over time
• Communicate behavioral progress to students, parents, and other school professionals

Operationally Defining a Behavior

To facilitate consistent and accurate data collection, educators must first develop an operational definition of the behavior—a clear and specific description of the behavior to be measured. Educators might collect data on the student’s engagement in a target behavior, replacement behavior, or both.

An operational definition should begin with a broad statement describing the behavior in terms that are:

• Observable—details what the behavior looks and sounds like, not what the student might be thinking or feeling
• Measurable—describes the behavior in such a way that it can be counted or timed
• Actively stated—specifies what the student does rather than what they fail to do

When operationally defining a behavior, it is also helpful to include specific examples and non-examples of the behavior. These additional details help clarify the definition by illustrating what constitutes a behavior when collecting data. Below are examples of operational definitions for a student’s target and replacement behaviors.

Systematic Direct Observation (SDO)

Educators can use a variety of methods to collect behavioral data in a systematic and consistent manner. One such method is systematic direct observation (SDO), which requires an observer to precisely measure how often a target or replacement behavior occurs or how long it lasts. Data can be captured using one of the following.

• Frequency or event recording—documents the number of times a behavior occurs within a given period (e.g., the number of times a student calls out during a one-hour lesson)
• Interval recording—documents whether a behavior occurred during brief periods or intervals (e.g., the number of one-minute intervals in which the student was out of their seat for any portion of time)
• Duration recording—documents how long a student engages in a behavior (e.g., time spent on task during independent work)
• Latency recording—documents the amount of time that elapses between when a direction or prompt is given and when the student begins the behavior (e.g., time between giving a student an assignment and the student beginning to work)

Note: Although educators can collect SDO data while teaching, doing so can be challenging and can interfere with instruction. Having a dedicated observer is recommended to capture the most accurate data.

Steps for Implementation

After identifying and operationally defining the behavior to be measured, use the steps below to collect SDO data.

1. Choose the appropriate method: Select a recording method that is appropriate for the nature of the behavior being measured, as noted below. Inappropriate choices (e.g., using duration recording for short, frequent behaviors) can result in misrepresentation of behavior patterns and misleading data.
   • Frequency or event recording—use for relatively brief behaviors with a clear beginning and end (e.g., shouting, hand raising, kicking)
   • Interval recording—use for behaviors that occur frequently or continuously (e.g., off-task behavior, academic engagement) or for when it is not feasible to count or time every instance of the behavior
   • Duration recording—use for sustained behaviors where the primary concern is the amount of time spent engaging in the behavior (e.g., crying episodes, attention to task, tantrums)
   • Latency recording—use for capturing the time between an instruction being given and the behavior beginning or ending (e.g., delayed transitions, delayed response to directions)
2. Create an observation schedule: Schedule observations during times when the behavior is expected to happen and for long enough to capture multiple instances of the behavior.
3. Select or develop a recording form: Make sure the documentation system aligns with the method selected in Step 1. Data can be documented on paper forms, computerized systems, or mobile apps.
4. Identify and train observer(s): One or more individuals (e.g., paraeducator, school psychologist) can conduct observations while the educator goes about usual instruction. Observers should be trained to understand the behavioral definition and the procedures for data collection.
5. Conduct observations and record data: Using the observation schedule developed in Step 2, the observer watches the student and records instances of the defined behavior throughout the observation session.
6. Organize the data: Organize in a way that supports interpretation and decision-making. For example:
   • Numerical data from multiple days or weeks can be entered into a simple table or chart for comparison.
   • Plotting the data on a line graph supports visual analysis of changes in the amount, severity, or consistency of the behavior over time.

Tips for Implementation

• To simplify the collection of data, try using one of these strategies:
  • Frequency data—Make tally marks on paper, use a small handheld counter, or put a handful of small items (e.g., pennies, paper clips) in one pocket and move one item to the other pocket each time the behavior occurs.
  • Interval data—Use a stopwatch, smartphone app, or vibrating wearable device to signal the end of an interval and prompt you to record data.
  • Duration or latency data—Use a digital stopwatch to collect data in minutes and seconds; using a wall clock or wristwatch will not be as precise.
• Consider using technology tools (e.g., smartphone timer, mobile tally counter apps, web-based form-building tool) to make SDO data collection more feasible in a busy classroom.
• If you are unsure whether the behavior occurred (e.g., the student walks behind a bookshelf and cannot be seen), do not attempt to guess what happened. Document only what can be directly seen or heard.
• Add brief written notes about the nature of the observed behavior alongside the numerical data to provide additional description or context.
• Students sometimes behave differently if they know they are being observed or when a new person is in the classroom. To reduce this reactivity, try observing other students at the same time, integrating the observer into the classroom at other times, and using discreet recording methods.
• Calculate rates, percentages, or averages as appropriate to allow comparisons of data across observation periods of differing lengths. The chart below provides helpful calculations for each SDO method.

Method	Calculation	Formula	Example
Frequency	Rate of the behavior	Rate = frequency count ÷ length of observation	25 callouts ÷ 1.5 hours = 16.7 callouts per hour
Interval	Percentage of intervals in which the behavior occurred	Percentage = number of intervals with behavior ÷ total number of intervals x 100	42 intervals with off-task behavior ÷ 60 total intervals x 100 = 70% of intervals with behavior
Duration	Percentage of time engaged in the behavior	Percentage = minutes engaged in the behavior ÷ total minutes observed x 100	6 min out of seat ÷ 20 min observed x 100 = 30% of time engaged in target behavior
Duration	Average length of behavioral instances	Average = sum of lengths ÷ number of instances	(2.2 min + 6.1 min + 3.5 min + 5.0 min) ÷ 4 instances = average length of 4.2 min
Latency	Average elapsed time	Average = sum of latency lengths ÷ number of instances	(23 sec + 90 sec + 49 sec) ÷ 3 instances = average elapsed time of 54 sec

Sample Recording Forms

For information about other data collection methods as well as graphing data, visit the following IRIS Case Study:

• Measuring Behavior