Kagan's Articles - FREE Kagan Articles

Research & Rationale

The Effects of RoundTable Consensus on the Social and Task Engagement of Students with or at Risk of Disabilities

Special Article

The Effects of RoundTable Consensus on the Social and Task Engagement of Students with or at Risk of Disabilities

Timothy G. Zgliczynski, Ph.D.
Abdullah Alshehri, Ph.D. BCBA

To cite this article: Zgliczynski, T., Alshehri, A., The Effects of RoundTable Consensus on the Social and Task Engagement of Students with or at Risk of Disabilities. Kagan Online Magazine, Issue #60. San Clemente, CA: Kagan Publishing. www.KaganOnline.com


The present study investigated the effects of Kagans’ cooperative learning structures on the social and task engagement of students with or at risk of a disability and their peers placed in a general education inclusive classroom. Three students with a low social profile (LSP) identified with or at risk of a disability and three typical peers identified with a high social profile (HSP) participated in this investigation. In general, LSP students showed more social engagement during the intervention than in baseline. In contrast, students with a HSP showed lower levels of social engagement during the intervention phase. In general, for the measure of task engagement, students showed more variability in their levels, although all students were more engaged in tasks when incentives were used. Lastly, all students reported they liked the intervention and it made them more focused on the task as well as connected to their peers.


The benefits of using the Kagan cooperative learning structures are readily visible when applied to any given lesson. Students instantly become engaged and participate at a level not typically observed in traditional whole-class instruction. While studies have explored the academic benefits of the Kagan Structures (Haydon et al., 2010; Maheady et al., 2006), few studies have examined the social and task engagement of students as they work within the Kagan Structures. Such examination is both timely and relevant as new standards emerge in order to address the social-emotional wellness of students as they attend school. Kagan and Kagan (2015) similarly addressed the social skill crisis in school as external environmental influences have caused both social and moral skills to decline.

The current social skills crisis is not limited to any one population of school-age child, however, students who are at-risk or identified with a disability have a greater likelihood of developing deficits in their social development (Bruininks, 1978). These students tend to be significantly less socially accepted than their peers. Furthermore, a longitudinal study conducted in grades three to six reported students with learning disabilities were viewed lower in social status consistently as they progressed through the formative years of their education (Estell et al., 2008). It is also not uncommon for students who struggle academically to engage in inappropriate behavior during instruction. Research has shown that disruptive behavior further leads to a decrease in levels of social acceptability, which compounds both academic progress and social/emotional development (Putnam, Markovchich, Johnson, & Johnson, 1996).

Cooperative learning presents an opportunity for students considered to be at risk and those with disabilities to practice academic and social skills deemed acceptable in the general education classroom. Cooperative learning has even been shown to improve social acceptance and friendships of such students (Slavin, 1989). Other research has also suggested at-risk students are viewed more favorably by peers when allowed to work together in cooperative groupings (Putnam et al., 1996). It is, therefore, reasonable to assume the Kagan Structures would yield similar, if not more profound, results. The basic principles of the Kagan Structures—(a) positive interdependence, (b) individual accountability, (c) equal participation, and (d) simultaneous interaction—all help to stabilize and regulate the cooperative learning environment and potential variables that can lead to social isolation. To date, few studies have thoroughly examined the potential social benefits of the Kagan Structures in a controlled experimental environment.

The purpose of this study was to investigate the effects of Kagan Cooperative Learning on the social and task engagement of students in an inclusion classroom. Specifically, one structure, RoundTable Consensus, was selected due to its compatibility with the learning outcome of academic vocabulary acquisition. On-task behavior, also known as task engagement, was previously defined by Haydon, Maheady, and Hunter (2010) as having one's eyes and/or marker on a required assignment and having eyes on peers while discussion material. Social engagement had been described by Harrell, Kamps, and Kravits (1997) as initiations that attempt to engage peers in mutual activity through movement or speech that include, but are not limited to actions of greeting, sharing, assisting, touching, calling a name, giving directions, or other social directed behavior. All measurements taken during the study were focused on these two variables and progress was determined by the frequency of their occurrence. Measurements were taken across various conditions in an attempt to isolate which variable make the biggest difference in student performance.


The study, which took place in an upper-elementary school in a suburb in the Northeast of the United States, was conducted across three different classrooms of students who received their social studies instruction from the same classroom teacher. All three classrooms were defined as consultant teacher model inclusion classrooms consisting of four to six students identified with disabilities. In addition, each class contained two to three additional students who were identified as being at-risk for gaps in academic and/or social development. During the student selection process, multiple teachers nominated students using the criteria from the Systematic Screening for Behavior Disorders (Walker & Severson, 1992). The criteria included such characteristics as low frequencies of speaking with other children, preferring to spend time alone, avoiding social situations, and being unresponsive to the social initiations of others. Once target students were identified, the teachers ranked them from least to most socially withdrawn. The top-ranking students (most socially withdrawn) were considered potential participants.

In the second phase of the selection process, each student in the three classrooms was given a three-question sociometric assessment (Leung & Silberling, 2006). The assessment asked students to name three classmates with whom they (a) were “best friends”; (b) most wanted in their small work groups; and (c) would most enjoy going to see a movie. If students were nominated as a first choice, they were awarded three points, two points for being the second choice, and one point for being the third choice. These points were totaled and the students who received the fewest points (the least amount of peer nominations) were considered potential target students for being at risk of having a low social profile (LSP). The results of the assessment were then paired with the teacher rankings and the students who scored low in student nominations and high in teacher ranking were selected for the study.

In addition to students with LSP, students were also selected from each class to be the “social stars” in the study if they were identified with a HSP. These students received a high number of peer nominations on the sociometric assessment and were also identified as having a high social profile (HSP) by the classroom teachers. In total, six students were selected to be target students in the study. Each of the three classrooms produced a student with a HSP and a student with LSP.  Out of the three students with LSP, only one, Brandon, was identified with a disability. All students with HSPs were typically developing students. Table 1 summarizes the target students across all three classrooms.

Table 1

Participant Summary Across Classrooms


The academic basis for the study was the acquisition and practice of content-related vocabulary in the three fourth grade social studies classes. For this purpose, each session utilized the Frayer graphic organizer as the means and a place for students to process their learning. Frayer models are advanced organizers that use four sections (definition, illustration, examples, and non-examples) to develop a holistic understanding of the word (see Figure 1). The classroom teacher had reported that all students had significant experience with the use of Frayer models and no additional training was required prior to the start of the study.

Figure 1. Frayer model for vocabulary instruction.

In addition to the Frayer model, the structure of RoundTable Consensus was utilized during certain phases of the study. According to Kagan and Kagan (2015), RoundTable Consensus meets the interpersonal functions of teambuilding, social skills, communication skills, and decision making. This structure, therefore, fit nicely with social skills focus of the study. All three classrooms included in the study were familiar with the Kagan Structures but had not used RoundTable Consensus before. For this reason, the researcher and classroom teacher held multiple training sessions where students were allowed to perform the duties of RoundTable Consensus, both with and without the use of Frayer models. Once students exhibited proficiency at using both these tools, the study was allowed to commence.

The methodology chosen for the study was a single subject design known as multiple-probe across participants. This design allowed for intermittent baseline data collection, with a minimum of three data points before an intervention phase can begin. For this study, there were three conditions. In condition A, baseline was collected after students had been subjected to traditional teacher-led instruction and unstructured group work. In condition B, teacher-led instruction was followed by group work using RoundTable Consensus. Since each lesson could be potentially different due to changing subject matter, it was decided that the classroom teacher would follow the explicit instruction lesson format (Archer & Hughes, 2010) in which there was an “I do, We do, and You do” phase. The “I do” lesson portion consisted of traditional teacher instruction. During this part of the lesson, the teacher taught using their typical lesson materials and introduced the target vocabulary word of the day. After the word had been contextually taught, the lesson transitioned to the “We do” phase. It was during this phase that students either entered into unstructured group work (first condition) or structured group work with RoundTable Consensus (second condition) in order to complete a Frayer model. Lastly, the students transitioned to the “I do” phase where they individually completed their own Frayer models based on the work their group had produced. The third condition, condition C, was identical to condition B but included the addition of an incentive.

Condition A - Baseline with Unstructured Group Work

For each condition, all students in the class were placed in Kagan groups. The two target students in each class (one student with a HSP and one with LSP) were placed within the same group in each of the three classes. During baseline data collection, the groups all partook in the teacher’s instructional delivery, which included the social studies word of the day. When the teacher moved to the “We do” phase of the lesson, the students were all given a blank copy of the Frayer model. They were then given 10 minutes to complete the graphic organizer to the best of their ability. Students were encouraged to work together with the other members of their group, however, no direction on how this should be done were given and no Kagan structure was suggested. If students or groups asked questions of the teacher, they were only given feedback related to the organizer and no additional direction on how the group should work to complete the task. Before the start of the following day’s lesson, students were all given another blank Frayer model and asked to complete it based on their learning from the previous day’s lesson. This allowed the teacher and the researchers to test for retention of learned content.

Condition B - Structured Group Work with RoundTable Consensus

After several sessions of baseline (three or more), the groups, one at a time, moved to condition B. Group arrangements were not changed during this time. Once the groups were properly trained in RoundTable Consensus, the structure was applied to several social studies vocabulary lessons. The lesson began in similar fashion to Condition A with the teacher delivering the content instruction for the day. Once this “I do” phase was complete, students were shown slides that reminded them of the steps needed to complete RoundTable Consensus. Another slide showed students which section of the Frayer model each numbered member of the group would be responsible for (i.e., 1 - definition, 2- illustration, 3- examples, 4- non-examples). The parts of the Frayer model assigned to each member of the group was counterbalanced and rotated with each lesson so each member of the Kagan group would be assigned all four sections equally.

Once students were aware of the condition and their responsibility, they were given a blank Frayer model and two minutes to complete just their assigned section. Kagan TimerTools were used to measure and display this time to students. After completing their own section, each table was handed a master Frayer model on a clipboard. Then each member of the group was given two minutes to present their section of the Frayer model to the group. The group then came to consensus by either accepting the student’s presentation or suggesting changes to it. Once consensus was met, the presenting student recorded the agreed-upon answer on the master Frayer model for the entire group. This process repeated until each member of the group presented their findings, came to consensus with the group, and recorded their findings on the master document in similar fashion. When time had expired, all members of the group were asked to copy the contents of the master Frayer model onto their individual Frayer model to help with retention. All Frayer models were collected as part of the “You do” phase of the lesson. Then students were asked to fill out a blank Frayer model about the vocabulary word on the subsequent day’s lesson just as they had in baseline.

Condition C - Structures Group Work with RoundTable Consensus and Incentive

Figure 2. Example of a Three Jars layout.

The third and final condition was identical to Condition B, however, an incentive was added to measure whether or not positive reinforcement would improve social performance within the groups. The researchers utilized the Three Jars system (Hawkins, Musti-Rao, Hughes, Berry, & McGuire, 2009), which was already established in the three classroom and highly valued by students. Three Jars (see Figure 2) is a group contingency reward system where randomization and the use of mystery motivators address the behavior and performance of all students in the classroom. As the name suggests, there are three glass jars in the classroom that are labeled “Who”, “What”, and “Win” respectively. The “Who” jar contained slips of paper with the names of all students in the three classrooms and the name of each team. Before beginning the “We do” phase of the lesson, the teacher would take the “Who” jar and select a single slip of paper from the jar. The paper may have had a student or team’s name, but the teacher did not tell the group who was selected, instead leaving that detail a mystery so that all students believed they could have been chosen. Next, the teacher selects from the “What” jar, which contains certain criteria the student/students chosen from the “Who” jar needs to exhibit in order for the whole class to win. For this study, the criterion placed in the “What” jar included: (a) staying on-task for 80% of the time, (b) implementing RoundTable Consensus with 90% or greater fidelity, (c) accurately scoring 90% or above on the graphic organizer (via rubric). Once the “Who” and the “What” were selected, the students were reminded of all the criterion that was in the “What” jar and allowed to begin their group work. The researchers, knowing the person/persons and criteria to look for, observed the individuals for either on-task behavior or RoundTable Consensus fidelity. If the graphic organizer accuracy was selected, then the researchers collected the Frayer models of the chosen student(s) and quickly checked them for accuracy.

Data Collection

Figure 3. Ricoh Theta camera used to record group conversations.

Video recordings served as the primary means for data collection. Ricoh Theta cameras were placed in the center of each table where students were working collaboratively. The Ricoh Theta camera (see Figure 3) is a wireless recording device capable of capturing 360-degree video. Recording in 360 degrees allowed the researchers to view the performance of the target students as if they were sitting in the middle of the table. The BORIS (Friard & Gamba, 2016) coding software was used to analyze the video content. For the task engagement measure, data was collected using 10-second whole intervals. In order for the interval to be scored as occurring, a participant had to be engaged throughout the entire 10-second interval. Social engagement data was collected in the BORIS software as well.

In addition to video data, social validity data was also collected in the form of exit interviews that were given to all target students in the study. The interviews were recorded on video and students were asked questions related to the social acceptability and validity of the study. The classroom teacher was also interviewed to measure his acceptability of the intervention and his overall experience of the study as a whole.

Results - Social Engagement

Overall results suggest (see Table 2) the inclusion of Kagan Structures improved the duration of social engagement in students with low social profiles. Large increases were observed for two of the three target students when switching from baseline to the use of RoundTable Consensus. Maria, the third target student, did decrease her duration in social engagement which was similar to the decrease seen among the three students with HSP. Despite the decrease, she was showing an upward trend in social engagement near the end of the intervention phase. Maria was also in the final class to enter intervention, so she remained in baseline for the longest amount in time which could have impacted her results. Maria’s teacher did state she had strong social skills but often chose not to use them during cooperative work.

The inclusion of an incentive also had an impact in social engagement. Both Nick and Maria saw large gains in social engagement. It is unclear why Brandon showed a decrease; however, one possibility is his desire to work as quickly and quietly as possible, which was a characteristic of the students with HSP in all three classrooms.

Table 2
Mean Scores for Total Duration (in seconds) for Social Engagement for LSP Students

Students with a HSP all displayed decreases in social engagement duration once RoundTable Consensus was introduced into the sessions. These results suggest participation was equalized once students were operating within a Kagan Structures. Students with a HSP, who tend to monopolize group work conversation, were not allowed to control the participation of each member and were required to share learning responsibilities. This, in turn, gave the students with low social profiles a chance to speak more often creating a negative correlation between the two groups. As students with a HSP were asked to speak less, students with low social profiles improved in their social engagement. This trend continued once the incentive was introduced. Duration of social engagement for students with HSP decreased once again. It is possible these students were intrinsically motivated by the idea of an incentive and made sure RoundTable Consensus was carried out with fidelity further equalizing the group members’ contributions.

Table 3
Mean Scores for Total Duration (in seconds) for Social Engagement for HSP Students

Results - Task Engagement

Results from on-task data yielded interesting results for students with low social profiles (see Table 4). Both Brandon and Nick had less time on task once RoundTable Consensus was introduced. Only Maria showed an increase in time on-task. One possible explanation for this phenomenon is the change in those students’ behavior from baseline to intervention. Brandon and Nick had low levels of social engagement during baseline and did not talk with their peers. Instead, these students spent time staring at their paper in writing, which characterizes on-task behavior. Once RoundTable Consensus was in place, Brandon and Nick no longer focused solely on their paper and instead chose to converse with their group. In contrast, Maria had a high score in baseline for social engagement and spent very little time attending to the task. Once RoundTable Consensus was introduced, she attended to the task and had less errant conversations with her peers. All three students did not show much of a change in on-task behavior when the incentive was added, although there were slight improvements. This suggests incentives had a mild effect on the on-task behavior of students with low social profiles.

Table 4
Participation Average Scores of Percentage for On-Task (Students with LSP)

On-task scores for the HSP group (see Table 5) were similar to those of the students with low social profiles. On-task scores slightly decreased for Connor and Maddie when RoundTable Consensus was introduced. Both students were accustomed to being in charge of the learning task during baseline and would often take command of the materials to make sure the job was done correctly. RoundTable Consensus made these students share the learning task which also led to a decrease in their on-task time. Robin did show an increase in on-task behavior once the intervention began, suggesting she chose to converse about off-task ideas when working without structure in baseline. All three students with a HSP showed an increase in on-task behavior when the Three Jars incentive package was applied. This suggest the use of Kagan Structures and incentive packages can improve on-task behavior in students who are academically and socially motivated.

Table 5
Participation Average Scores of Percentage for On-Task (Students with HSP)

Student/Teacher Interviews

The qualitative data for students’ perception were more consistent across participants. All six participants liked and learned through the Kagan Structure better than without it. This is consistent with previous investigation such as Alshehri (2017) and Haydon et al. (2010). Students expressed that they favored the structure because everyone in the group is divided (i.e. structured) and students are given turns. Students also agreed group work made them connect and understand their peers better.

All students also agreed that they felt more focused when the structure was in use and some thought that the structuring is what made them focus. Lastly, all students showed a preference for having the three jars incentives in place as that made them work harder on the task.


The results of the study show a positive functional relationship between the use of Kagan Structures on the social engagement and on-task behavior of students with low social profiles and those with higher social profiles. In addition, several of the participants in the study were diagnosed with a disability or were at-risk for academic deficits. It is highly possible that Kagan Structures could be useful in promoting inclusionary practices and leveraging participation across classrooms that contain students of varied social and academic competencies. There some limitations to the study. Academic data was collected but not analyzed. Additional research is needed to examine whether or not correlations exist between increased social engagement, on-task behavior, and academic outcomes. It is also important to note the small sample size and the fact that all three classrooms were taught by the same social studies teacher. Future studies may want to examine larger samples across more varied settings and subject matter. In addition, the use of different cooperative learning structures may yield different results.

Despite these limitations, educators should take these results as validation for using the Kagan Structures in their classrooms. The potential benefits for students with low social profiles, and all students in general, are worthy of time and investigation. Every classroom, regardless of makeup, consists of students with varied social competencies. The Kagan Structures can offer a safe, risk-free environment where students can develop their social skills, equally contribute in meaningful ways, and work toward a common goal. Inevitably, these skills can translate to success over the span of a lifetime.


Alshehri, A. (2017). Structured versus unstructured cooperative learning for the social interaction of students with disabilities: effects and students’ acceptance. Unpublished manuscript. Department of Learning and Instruction. University at Buffalo. Buffalo, NY, USA.  

Archer, A.L., & Hughes, C.A. (2011). Explicit instruction: effective and efficient teaching. New York: Guildford Press.

Bruininks, V. L. (1978). Actual and perceived peer status of learning-disabled students in mainstream programs. The Journal of Special Education, 12(1), 51-58. doi:10.1177/002246697801200109

Estell, D. B., Jones, M. H., Pearl, R., Van Acker, R., Farmer, T. W., & Rodkin, P. C. (2008). Peer groups, popularity, and social preference: trajectories of social functioning among students with and without learning disabilities. Journal of Learning Disabilities, 41(1), 5- 14. doi:10.1177/0022219407310993

Friard, O., & Gamba, M. (2016). BORIS: a free, versatile open-source event-logging software for video/audio coding and live observations. Methods in Ecology and Evolution, 7(11), 1325-1330. doi:doi:10.1111/2041-210X.12584

Harrell, L. G., Kamps, D., & Kravits, T. (1997). The Effects of Peer Networks on Social—Communicative Behaviors for Students with Autism. Focus on Autism and Other Developmental Disabilities, 12(4), 241–256. doi:10.1177/108835769701200406

Haydon, T., Maheady, L., & Hunter, W. (2010). Effects of numbered heads together on the daily quiz scores and on-task behavior of students with disabilities. Journal of Behavioral Education, 19(3), 222-238. doi:10.1007/s10864-010-9108-3

Hawkins, R., Musti-Rao, S., Hughes, C., Berry, L., & McGuire, L. (2009). Applying a randomized interdependent group contingency component to classwide peer tutoring for multiplication fact fluency. Journal of Behavioral Education. 18(4):300-318. DOI: 10.1007/s10864-009-9093-6

Kagan, S., & Kagan, M. (2015). Kagan Cooperative Learning. San Clemente, California: Kagan Publishing.

Maheady, L., Michielli-Pendl, J., Harper, G. F., & Mallette, B. (2006). The effects of numbered heads together with and without an incentive package on the science test performance of a diverse group of sixth graders. Journal of Behavioral Education, 15(1), 24-38.

Putnam, J., Markovchick, K., Johnson, D. W., & Johnson, R. T. (1996). Cooperative learning and peer acceptance of students with learning disabilities. The Journal of Social Psychology, 136(6), 741-752. doi:10.1080/00224545.1996.9712250

Slavin, R. E. (1989). Research on cooperative learning: consensus and controversy (Vol. 47, pp. 52). Washington, D.C: Association for Supervision and Curriculum Development.

Walker, H. M., & Severson, H. H. (1992). Systematic Screening for Behavior Disorders (SSBD) (2nd ed.). Longmont, CO: Sopris West.