Assessing the classroom-wide impact of a behavioral program for disruptive students: effects on disruptive students, classmates, and teachers
- Primary schools
- Student learning
- Tailored instruction
School-based social and emotional learning programs can be a promising way to reduce student disruptiveness in school. Researchers evaluated the impact of Chile’s national Skills For Life (SFL) program for disruptive students on their behavior and academic outcomes. Researchers found that the SFL program did not impact eligible students’ learning outcomes.
In school, one student’s disruptive behavior can impede the learning of all the students in the classroom. Specifically, being exposed to a disruptive peer can reduce classmates’ test scores. As a result, many teachers devote substantial class time to discipline. School-based mental health programs are a commonly used strategy to reduce students’ disruptiveness. These programs often focus on social and emotional learning—the process through which children acquire the skills to recognize and manage their emotions, set and achieve positive goals, and handle interpersonal situations effectively. They do so by trying to enhance children’s self-awareness, self-management, and social awareness, but research has been mostly limited to evaluating programs closely monitored by researchers and in high-income country settings. How does a large government-run social and emotional learning program targeted to select disruptive students in Chile impact student behavior and academic outcomes?
Context of the evaluation
ADHD, a disorder correlated with behavior problems, is more prevalent in Chile than in countries like the US, France, or Italy, which may exacerbate the problem of disruption in schools. Chile’s nationwide social and emotional learning program, Skills for Life (SFL), leverages cognitive and behavioral therapy for disruptive students in the second grade. Since 1998, SFL has worked with around 1,000,000 children, making it the fifth-largest school-based mental health program in the world. In 2001, SFL became a nationwide policy, and as of 2010, is implemented in 1,637 publicly-funded elementary schools—twenty percent of all elementary schools in Chile.
The classes invited to participate in the evaluation had an average of 33.2 students. The students generally came from low income, more disadvantaged groups. Specifically, about three-fourths of the students lived below the social security program threshold and about one-third of the students were born to teenage mothers, which was more than twice the corresponding proportion in Chile.
To identify students eligible to participate in the SFL program, teachers evaluated each student based on his/her ability to accept authority, attention and focus, activity level, social skills, interest in school, emotional maturity, overall disruptiveness, and academic ability. Based on scores across these metrics, about fifteen to twenty percent of all first-grade students are eligible for SFL each year. Eligible students for this evaluation were more likely to be male, to not live with their father, and to have parents with only 8.5 years of education. On average, eligible students had lower self-control and self-esteem, fewer and more disruptive friends, and lower test scores before the start of the program.
Details of the intervention
Researchers conducted a randomized evaluation to test the impact of Chile’s Skills For Life program for second graders with conduct problems on the behavioral and academic outcomes of all students in the classroom.
Researchers randomly assigned 172 second grade classes with 5,704 students to participate in the SFL program at varying times throughout the 2015 school year. In schools with enough eligible students to form two independent workshop groups, researchers conducted a lottery within the schools to randomly assign one class to receive the program in the first semester of 2015. The other class would receive the program in the second semester. For the remaining schools, randomization took place within municipalities. Altogether, 89 classes received the program during the first semester, while 83 classes received it during the second semester and served as the comparison group for this evaluation. Eligible students in these classes were then invited to enroll in a workshop.
SFL workshops consisted of ten two-hour group sessions that took place during the school day once a week over the course of one semester. The workshops were typically led by a pair of government-hired psychologists and social workers or teachers, and each group had six to twelve participating students. These sessions were activity-based, involved games and role play, and made use of cognitive behavioral therapy techniques which helped students think through their actions and manage their behavior. During sessions, enrolled students left their regular classroom, while their classmates remained and continued with the normal schedule. As an incentive for participation and good behavior in group sessions, students sometimes received a candy reward.
Researchers used data from JUNAEB, the division of the Department of Education that manages the SFL program, on student eligibility, student socio-economic background, student disruptiveness, and program implementation. Researchers also administered questionnaires to students to gather information on their social wellbeing and hard skills, like Spanish and math, as well as to teachers to gather their view of the classroom environment. Researchers also visited each of the classrooms for half a day.
Results and policy lessons
The SFL workshops had no measurable impact on eligible students’ socio-emotional skills, behavior, or academic outcomes. Evidence suggests, however, that SFL may increase students’ disruptiveness in the classes that have at least one very disruptive student.
Social and emotional learning: The SFL workshops had no measurable effect on the emotional stability of students who were eligible to participate in the program, as measured by student happiness, self-esteem, and self-control scores. The workshops also did not change the emotional stability of students who were not eligible to participate in the program.
Social networks and academic outcomes: The workshops had no measurable effect on the friendships of students eligible for the SFL workshops. Additionally, the SFL workshops did not change the academic outcomes—as measured by the proportion of school days missed, Spanish test scores, and math test scores—of students who were eligible to participate in the program or those who were ineligible for the program.
Levels of disruptiveness: The program did not change the disruptiveness of students eligible for the SFL workshop. However, in classes with at least one very disruptive eligible student (defined as students who scored in the 90th percentile of eligible students’ disruptiveness at the beginning of the study), the program strongly increases the disruptiveness of all other students, as well as teachers’ and enumerators’ ratings of the overall disruptiveness of the class. Evidence suggests that this may be because the program increases the friendship connections between very disruptive and other students, and it is possible that very disruptive students may then have a negative influence on other eligible students.
Cost of program: JUNAEB did not have an estimate of the total cost of the program due to variation in costs between implementing teams, so researchers estimated that SFL cost about US$200 per student that received the program, which translates to about 15 percent of what the government spends on the average student per year.
These results seemed to contradict a large literature that has found social and emotional learning programs implemented under researchers’ close supervision in high-income countries to be very successful. Researchers posited that this might have been because SFL was a large government-run program. Moreover, since ADHD, a disorder correlated with conduct problems is highly prevalent in Chile, SFL may be faced with a harder-to-treat population than other social and emotional learning programs, which could also explain why it does not seem to produce strong effects. In the end, the government did not change the program implementation because of this research.