Positive Behavior Support for a Child With Inattentive Behavior in a Japanese Regular Classroom

The Japanese School Context

We had been participating in Kobe City’s special education project of supporting children with special needs as an assistant teacher and as a consultant when we met Riku. The assistant teacher was also a graduate student studying applied behavior analysis, and the consultant was also a university professor teaching applied behavior analysis. Riku was a first-grade boy enrolled in a regular classroom in Kobe City’s public elementary school. The class he belonged to consisted of 36 students, which is a typical class size in urban areas of Japan. One female classroom teacher who was in her 2nd year of teaching was in full charge of teaching and managing the whole class. The assistant teacher was present only for 1 to 2 hr per week, supporting several children who needed help in Riku’s classroom, before we started any assessments or support programs for Riku. Although the number of supports and projects in supporting children with special needs is increasing, only few Japanese schools have outside personnel or additional teachers working in regular classrooms even for a limited amount of time. Classrooms had typical Japanese arrangement, with student desks set in rows facing the front and having very little space left. The predominant way of teaching and learning in such class and classrooms is through instructions directed to the whole class. Riku was referred to us by the classroom teacher as having difficulty attending to and/or responding to such whole-class–directed teacher demands.

Functional Behavioral Assessment

According to the classroom teacher’s report and the assistant teacher’s informal observations, Riku exhibited inattentive behavior especially during math classes. Moreover, according to the assistant teacher’s informal observations, apparent consequences such as teacher attention or engagement in a preferred activity were absent. Therefore, we decided to conduct an antecedent-based FBA during math classes.

We implemented an antecedent-based FBA using the following procedures. For each session, which was composed of a 45-min-long math class, we recorded each teacher demand directed to the whole class and coded Riku’s behavior in response. We audiotaped all whole-class teacher demands for accuracy and categorized them into four types. Examples of the four demand types were “Please open page 65” (Direction), “In this picture, how many cars are parked in the parking lot?” (Question), “how to compute 9 + 4” (Explanation), and “solving problems on pages 67 and 68” (Task). We coded Riku’s response behaviors as on-task when he was (a) complying to “Directions,” (b) raising his hand to answer “Questions,” (c) engaging in classroom “Tasks,” or (d) behaving in ways related to working out answers to “Questions” and listening to “Explanations,” such as looking forward, looking at the teacher’s face, or looking at academic materials. We coded all other behaviors that occurred after teacher demands as inattentive behavior. Some typical examples of inattentive behavior were fiddling and looking elsewhere. In regards to “Tasks,” we also considered nonengagement for more than 1 min or not being able to keep up with the class pace when checking answer as inattentive behavior. During this antecedent-based FBA sessions, which also served as Baseline 1 sessions in Figure 1, Riku’s percentages of on-task behavior under differing types of teacher demands were as follows; Directions (28%), Questions (50%), Explanations (40%), and Tasks (76%). We also computed Riku’s percentages of on-task behavior during the first half and the latter half of the class during the FBA sessions. Riku had higher percentages of on-task behavior during the first half of the class (mean of 46%) compared to the latter half of the class (mean of 38%).

OPEN IN VIEWER

Figure 1.

Riku’s percentage of on-task behavior

Apart from this antecedent-based FBA sessions, in additional four sessions, we also assessed the percentage of on-task behavior of all children in the class including Riku using an interval recording procedure. Using a stopwatch, we observed two children for 5 s and judged whether on-task behavior was present or not among the two, and recorded in the next 5 s. We repeated this procedure until all children were assessed eight times during one session. Riku’s mean on-task behavior assessed through an interval recording procedure was 62%, compared to the class average of 88%.

We collected interobserver reliability data for 13% of the Baseline 1 sessions and for 20% of the interval recording sessions. We calculated the reliability rate for on-task behaviors (%) during Baseline 1 sessions by dividing the number of agreements of on-task behaviors by the number of teacher demands given and multiplying it by 100. Similarly, we calculated the reliability rate for on-task behaviors during interval recording sessions by dividing the number of intervals of on-task behaviors agreed by the whole number of intervals and multiplying it by 100. Interobserver reliability rates were 90.5% and 89.5%, respectively.

Assessment results revealed that Riku’s percentage of on-task behaviors was low, especially in response to auditory-based teacher demands such as Directions, Questions, and Explanations compared to teacher demands that included more visual cues, that is, Tasks. Riku’s low on-task rates were consistently apparent in the additional interval recording assessment, in which his peers engaged in high rates of on-task behavior. These results approved the need for a support program for Riku. Another finding of Riku, doing better during the first half of the class when content of demands were easier compared to the second half, raised a question of whether he was capable of performing on-task behaviors from an academic perspective. Did he have the academic abilities necessary?

Positive Behavior Support

In answering the question that arose, and in order to clarify whether common supports are effective, we implemented individual support (IS) and verbal praise (VP) measures. The assistant teacher gave individual support of repeating, rephrasing, or explaining the classroom teacher’s demands often along with behaviors such as pointing at materials, patting on Riku’s shoulders, or calling Riku’s name to gain attention when Riku failed to follow the classroom teacher’s whole-class–directed demands. In addition, the assistant teacher gave verbal praise when Riku exhibited on-task behavior after individual support. Figure 1 indicates that during the IS and VP phase, Riku’s on-task rates increased by 30% compared to that with the Baseline 1 phase. The common supports of IS and VP were not only effective in raising Riku’s on-task rates but it also suggested that Riku was capable of performing academically.

Considering the school context, we preferred an effective support program that could be easily implemented by the classroom teacher. With assessment results and IS and VP results, the assistant teacher and the classroom teacher talked and decided on implementing Preceding Support–whole class (PS-wc) and Preceding Support–individual (PS-i) measures. The classroom teacher gave PS-wc of gaining the attention of the entire class by saying, “Everybody, I want you to look forward,” or waiting for a few seconds for the children’s attention before giving whole-class–directed demands. The classroom teacher also gave PS-i of gaining Riku’s attention individually by statements such as “We are here right now,” often along with the behavior of pointing at his materials prior to giving whole-class–directed demands. The classroom teacher gave PS-wc for an average of 2.0 times per session and PS-i for an average of 2.4 times per session. Figure 1 indicates that Riku’s mean on-task rate was 74% during this PS-wc and PS-i phase, showing a higher rate compared to those in Baseline 1 (40%) and Baseline 2 (51%). During a follow-up phase conducted 1 to 3 months after this phase, Riku’s on-task rate still remained 10% to 20% above those in the baseline phases. The teacher’s ratings on a Likert-type scale consisting of six questions showed high acceptability and the effectiveness of PS-wc and PS-i (mean item score of 4.7, range 4–6 on a 6-point scale).

Conclusion

We see first- or second-grade students who exhibit nondisruptive problem behaviors such as inattention, nonengagement, or quietly playing with their hands/materials in many regular classrooms in Japan. However, because of its large class size and its group-oriented characteristic, together with the quiet feature of these students’ behaviors, it is not uncommon for many of these students to remain unsupported. This is an unfavorable trend because these students are at higher risks of academically underachieving or of developing more serious problem behaviors in later grades. Our case targeting a first-grade boy’s inattention suggests that certain teacher-friendly supports such as PS-wc and PS-i can help and increase such students’ engagement in the classroom. Applied behavior analysis’ three-term contingency and the collection of data are also very easy to understand and are helpful tools for classroom teachers who are facing difficulty in instructing students with problem behaviors.

Our attempt of applying FBA and FBA-based positive behavior support to a nondisruptive problem behavior of inattention in a regular classroom suggested efficacy to some extent. However, studies of FBA and FBA-based support applied to similar behaviors and to more participants are necessary. The application of FBA has been expanding to new populations and to various behaviors over the years and is continuing. In addition to its use for challenging behaviors, its use in early interventions or for the promotion of desirable behaviors may be one direction to be investigated with more emphasis in the future.