BA420 Unit 3 Discontinuous Measurement Paper Example

Twenty-Two Percent or Forty-One: A BA420 Unit 3 Discontinuous Measurement Paper

[Student Name]

Purdue University Global

BA420: Methods of Behavior Analysis

Unit 3 Discontinuous Measurement Paper

[Instructor Name]

[Date]

The writer, the school, the student and every figure are composite and illustrative; the article and books cited are real.

The Problem

Jaylen's teacher asked me a practical question in October about how to track his behavior herself. Jaylen, a sixth grader, spends a noticeable part of his math period out of his seat, and his teacher wanted to track it every day. I can time his behavior continuously during math, using a stopwatch app that logs when each episode begins and ends. His teacher cannot; she is teaching. She needed a reliable, simple method she could use while teaching, which meant a discontinuous one, recording behavior at intervals rather than continuously. The question was which discontinuous method would give her numbers close enough to the truth to guide decisions about Jaylen's plan.

Cooper et al. (2020) describe three common discontinuous methods. In partial-interval recording, the session is divided into intervals, and an interval is scored if the behavior occurs at any time during it. In whole-interval recording, an interval is scored only if the behavior occurs throughout the entire interval. In momentary time sampling, the observer looks at the end of each interval and scores whether the behavior is occurring at that moment.

To choose among them, the district behavior analyst asked me to record one fifty-minute period in four ways at once, using a video of the class taken with written parent consent for that purpose and deleted after scoring.

Four Numbers for One Period

Continuous timing showed that Jaylen was out of his seat for eleven minutes of the fifty, or 22 percent of the period. With ten-second intervals, partial-interval recording scored 41 percent of intervals, whole-interval recording scored 12 percent, and momentary time sampling scored 23 percent. With sixty-second intervals, partial-interval recording rose to 68 percent, while momentary time sampling stayed close, at 25 percent.

Partial-interval recording nearly doubled Jaylen's time out of seat, and with longer intervals it tripled it.

The errors follow directly from the rules. Partial-interval recording counts an entire interval if the behavior happens for even a second within it, so short episodes spread across many intervals inflate the estimate, and longer intervals inflate it more. Whole-interval recording misses any episode that does not fill an interval, so it underestimates. Momentary time sampling neither inflates nor deflates systematically; it samples the behavior at moments, so with enough samples its estimate approaches the true proportion of time.

What the Research Shows

Powell et al. (1975) compared time-sample measures against continuous recording of the same behavior. They found that momentary time sampling produced estimates closer to the actual duration than partial-interval recording, which tended to overestimate, and whole-interval recording, which tended to underestimate. The interval length mattered too, in their data as in mine, with shorter intervals generally producing more accurate estimates. Jaylen's data reproduce their findings almost exactly, nearly fifty years later.

Choosing a Method

After reviewing the comparison, the team chose momentary time sampling with thirty-second intervals for the teacher's daily data. A vibrating timer on her wrist signals each moment, and she glances at Jaylen and makes a mark on a small grid clipped to her lesson plan. The whole process takes about a second each time, which she said she barely notices after the first week. Thirty-second intervals were a compromise: ten-second intervals were most accurate but impossible while teaching, while sixty-second intervals gave too few samples in a period to be stable. A check with the video showed thirty-second momentary time sampling produced 24 percent against the continuous 22 percent.

The team also built in a safeguard. Once a week, I time Jaylen continuously during the same period the teacher samples, so that the two measures can be compared. If momentary time sampling begins to diverge from continuous timing by more than a few percentage points, for example because Jaylen's episodes become shorter and more frequent, the team will reconsider.

Training the Teacher

A feasible method still has to be learned. I trained Jaylen's teacher in three short sessions after school. First, we watched five minutes of the class video together, and I explained the definition and the moment-of-glance rule: score only what Jaylen is doing at the instant the timer vibrates, not what he did a few seconds earlier. Second, she scored ten minutes of video on her own, and we compared her marks with mine; we agreed on 85 percent of moments, with most disagreements coming from moments when Jaylen was half out of his chair. We clarified the definition for that case. Third, she practiced live for two days while I scored alongside her, and our agreement rose above 90 percent. Only then did her data begin to count.

What the Numbers Changed

The choice of method mattered for Jaylen. Before the comparison, the teacher had been keeping a rough partial-interval record with five-minute intervals, which showed him out of his seat in nearly every interval. That record had supported a proposal to move him to a different classroom. The continuous and momentary data told a different story: he was out of his seat about a fifth of the time, mostly during independent practice, and rarely during direct instruction. The team dropped the classroom change and instead added two scheduled movement breaks during independent practice. Within a month, his time out of seat fell to about 12 percent.

When Partial-Interval Recording Makes Sense

The comparison does not mean partial-interval recording is never useful. Kazdin (2011) notes that interval methods can be appropriate when the goal is to detect whether a behavior occurs at all within periods, or when behavior is too frequent or brief to count. For a behavior the team wants to eliminate, such as aggression, partial-interval recording's tendency to overestimate errs on the side of caution. The method should match the question and the behavior's characteristics, and whoever chooses it should know in which direction it is likely to err.

Conclusion

The same period of Jaylen's behavior produced estimates from 12 to 68 percent depending on the method and interval length, while the true figure was 22 percent. Powell and colleagues' research explains why: partial-interval recording overestimates, whole-interval recording underestimates and momentary time sampling approximates duration best. The team's choice, thirty-second momentary time sampling with weekly continuous checks, gives a busy teacher a method that is both feasible and close to the truth.

References

Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.). Pearson.

Kazdin, A. E. (2011). Single-case research designs: Methods for clinical and applied settings (2nd ed.). Oxford University Press.

Powell, J., Martindale, A., & Kulp, S. (1975). An evaluation of time-sample measures of behavior. Journal of Applied Behavior Analysis, 8(4), 463-469. https://doi.org/10.1901/jaba.1975.8-463