When and How Should Behavioral Data Be Captured in the Classroom?
Recordings about classroom behavior are most accurate when done right after the behavior occurs. Recording at the end of a lesson is next most accurate and recording 3 hours later is the least accurate. These statements seem intuitive. However, a careful study described in a recent paper* found that many factors affect these differences in accuracy.
Methods: The study involved special educators recording about a single target behavior such as crying, eye poking, and “talk-outs”. The educators noted the occurrence of the behavior either 1) right after it happened, 2) after the activity in which it happened finished, or 3) a few hours later. Recordings were made on paper forms. Four teachers participated and each recorded about two of their middle/junior high students. They received training on recording and treatment fidelity. Their recordings were compared to those made by one of the investigators using videotaped observations. The investigators randomly varied under which of the three conditions the teacher recorded for each of two recordings per day. Teachers recorded data on student performance across 15 days with 10 sessions for each condition.
Results: All four teachers met the necessary 85% agreement level for their data to be considered accurate for at least one student during either Conditions 1 or 2. Three teachers did not record accurate data for their other student during these same conditions. Each had higher levels of agreement among some conditions than others. Not all teachers consistently recorded accurate data for both of their students across conditions. An important factor impacting accuracy was the overall frequency of the behavior. Other factors that also had an impact include: treatment fidelity, medication or other treatment changes, and the salience of the behavior during the instruction.
Conclusions: The authors conclude: “Although immediate direct observation recording is espoused as the most accurate data recording method, teachers did not consider it the most feasible particularly because of the barriers they face when using this data recording method. Prior research provided evidence to suggest special educators can engage in accurate delayed recording; however, the delay between behavior occurrence and recording should be 30 min or less…Future research should consider choosing target behaviors that are less prone to salience and do not pose difficulties for teachers to maintain high levels of treatment integrity and reliable data recording (e.g., moderate frequency behaviors).”
Comment: The study did not include a condition with electronic data capture. SymTrend screens (above) can be used to capture data in the moment (left) or at the end of the lesson/3 hours later (right). When done in the moment the teacher taps the screen each time a behavior occurs and the screen then refreshes. This cannot be done as the teacher is doing a lesson; a second observer would enter the data. The teacher could use a “clicker” to record each time a behavior occurred and then enter the total on the screen on the right after the lesson is over. Teachers who have used SymTrend preferred the accuracy of data recording by paraprofessionals not immediately involved in the lesson. Transcription errors sometimes occurred when copying from paper tally sheets to screens like the one on the right. Transcription from clickers provided intermediate accuracy, when available.
* Jasper, AD and Taber Doughty, T. (2014) Special educators and data recording: What’s delayed recording got to do with it? Focus on Autism and Other Developmental Disabilities. Published online 12 August 2014.