Abstract
James is a seventh grader at Harlan Middle School. James receives special education services for his learning disability. In addition, his school uses a positive behavior interventions and supports (PBIS) framework to address the behavioral needs of all the students in the school. Per James’ individualized education program (IEP), he is provided with various accommodations and specialized instruction in reading and mathematics. However, his IEP does not specify any behavioral supports. This is unfortunate as James struggles throughout his school day with staying on task in the classroom, following teacher directions, and completing assignments on time. He has been sent to the office three times in the first quarter for various inappropriate behaviors during math class. Faculty and staff are frustrated with James as he is not responding to Tier 1 behavioral supports. The PBIS team determines that James needs to receive Tier 2 supports. The team knows that it will need to monitor James’ responsiveness to the Tier 2 intervention in order to determine if a more intensive intervention is effective but wonders what to do if James is responsive to the intervention. In other words, if the Tier 2 intervention works, should James just go back to the Tier 1 supports, or should the Tier 2 supports be maintained indefinitely?
Behavior problems are not limited to students of a specific gender, socioeconomic status, age, or ability. Rather, students who misbehave represent a range of backgrounds, including those with disabilities (e.g., learning disabilities, speech and language impairments, emotional and behavioral disorders, autism spectrum disorder) and without disabilities. Further, behavior problems are likely to be present in both general and special education classrooms (Forness, Kim, & Walker, 2012). Within the classroom, students with persistent challenging behavior may display attention difficulties, noncompliance, and low self-control, among other undesirable behaviors. Given students with behavior problems also may struggle academically, it is crucial to provide targeted (Tier 2) and intensive (Tier 3) behavioral supports necessary to prevent high rates of school failure, dropout, grade retention, and suspension.
To this end, thousands of schools across the country have adopted PBIS as a way to prevent and respond to challenging behavior. PBIS is a federally endorsed, three-tiered approach for promoting positive behaviors and implementing a continuum of data-driven interventions to students (Individuals with Disabilities Education Act, 2004). All students who attend schools that employ PBIS receive Tier 1 supports, which include (a) teaching universal, prosocial behaviors to prevent future problems; (b) implementing evidence-based classroom management strategies; and (c) acknowledging students when they display positive behaviors. Students who fail to respond to Tier 1 supports move to Tier 2, which are typically small-group, evidence-based behavioral interventions and supports (Wehby & Kern, 2014), although Tier 2 can be delivered at the individual level. Finally, in Tier 3, students with high-risk behaviors who fail to respond to Tier 1 and Tier 2 supports (<5% of school population) receive highly individualized, intense behavioral interventions, which often are designed based on data collected from a functional behavior assessment (Debnam, Pas, & Bradshaw, 2012). Ideally, data are used to facilitate a fluid process of moving students between tiers based on response to intervention.
The intervention must be faded systematically by gradually reducing and eventually eliminating key elements, such as (a) goals, (b) feedback, (c) reinforcement, (d) dosage, or (e) other intervention components.
Using student data to determine student responsiveness and make subsequent intervention decisions is often referred to as data-based individualization (Fuchs, Fuchs, & Malone, 2017). The type of data collected varies by intervention type and behavior targeted for improvement. Data can include office discipline referrals, attendance records, direct behavior ratings, systematic direct observation, and intervention-based measures (Bruhn, McDaniel, Rila, & Estrapala, 2018). Although much has been written about intensifying interventions for nonresponders (Fuchs et al., 2017; Wehby & Kern, 2014), less information is available to teachers for reducing the intensity of interventions when students experience success (i.e., responders). Consequently, many teachers may not have the skills or knowledge to systematically fade intervention supports based on student data (Dunn, Airola, Lo, & Garrison, 2012). Instead, teachers frequently end up stopping an intervention “cold turkey” once the student demonstrates improved behavior. This often leads the student to revert to preintervention levels of behavior, which defeats the purpose of teaching and reinforcing desired behaviors (Cooper, Heron, & Heward, 2007). Teachers can avoid this relapse by systematically fading intervention components.
Programming for Maintenance Through Systematic Fading
Ensuring that newly acquired behavioral skills continue after an intervention has ended (i.e., maintenance) is the ultimate goal of behavioral interventions (Baer, Wolf, & Risley, 1968). But no matter how effective an intervention is, positive behavior is unlikely to sustain if the intervention is stopped abruptly after the student experiences success (Rusch & Kazdin, 1981). The overarching theory is not that behavior is “cured” and intervention can be stopped “cold turkey,” but instead, the intervention must be faded systematically by gradually reducing and eventually eliminating key elements, such as (a) goals, (b) feedback, (c) reinforcement, (d) dosage, or (e) other intervention components. The process of systematic fading involves removing or reducing one intervention component at a time while continuing to monitor student data (Cooper et al., 2007). Slowly removing intervention components and monitoring student responses enables a practitioner to analyze the effects of intervention components and increase the likelihood that positive behavioral change will maintain over time (Rusch & Kazdin, 1981). Because initial success does not guarantee long-term success, a gradual and systematic approach to individualizing interventions based on data about student response is the best way to maintain behavioral change.
Determining how to adapt an intervention to ensure long-term student success is complex. There are no hard and fast rules as to which intervention component should be adapted first or which should be adapted last. There are, however, a few principles teachers should consider before acting. First, systematic fading refers to a plan for slowly adapting or reducing only one component at a time, like dosage (e.g., self-monitoring every other day instead of every day). If more than one component is faded and behavior begins to regress, knowing which component was working best for the student may be unclear. Instead, teachers should make their best judgment about one component to adapt first based on individual student characteristics and contextual fit. For instance, feedback might need to be faded later in the process if the student thrives on teacher attention. Related to contextual fit, it is important to consider which components may be sustained for long-term use (e.g., intermittent praise) versus components that are difficult to maintain every day.
Second, systematic fading for responders is predicated on analyzing student data and making adaptations accordingly. Just as data collection and analysis are ongoing during implementation of an intervention, these same procedures need to continue throughout the fading process. By continuing to monitor student progress, it is possible to see if making gradual adaptations to interventions is effective or if the fading process is resulting in relapses in negative behavior. If this happens, best practice is to return to the original intervention plan until high and stable levels of positive behavior are demonstrated. Then, fading a different intervention component is recommended. This iterative process of reducing or removing intervention components continues until all intervention components are faded completely, with the ultimate goal of transitioning students back to only Tier 1 supports.
Teachers should make their best judgment about one component to adapt first based on individual student characteristics and contextual fit.
How to Fade Tier 2 Components
As noted previously, students with and without disabilities who demonstrate challenging behaviors within a Tier 1 PBIS framework can receive Tier 2 supports. Typically, Tier 2 interventions should not require a lot of time or resources because they are designed to serve up to 15% of the school population. Common Tier 2 interventions include check-in/check-out (CICO; Crone, Hawken, & Horner, 2010), small-group social-skills instruction, and self-regulation skill development. CICO involves students checking in with a mentor at the beginning and end of the day, receiving teacher feedback via a CICO report card (sometimes referred to as a daily progress report), monitoring their progress toward goals, and bringing home the daily progress report to their parents. In social-skills interventions, students meet with a school staff member, such as a counselor, teacher, or paraeducator, usually one to three times per week in small groups of three to five students who have similar issues. Instructors may use a prepackaged curriculum to teach, model, practice, and reinforce social skills, such as cooperation and listening (Rodriguez, Loman & Borgmeier, 2016). Self-regulation interventions often involve multiple components, such as goal setting and self-monitoring. Goal setting includes teaching students to (a) identify problem behaviors, (b) develop a measurable and attainable goal for improving behavior, and (c) execute a plan for monitoring their performance toward the goal (Bruhn, McDaniel, Fernando, & Troughton, 2017). Self-monitoring is the act of students thinking about their behaviors and then recording the presence or absence of those behaviors. Although there are other possible Tier 2 interventions, these common interventions serve as an excellent model for demonstrating how the intervention components of (a) goals, (b) feedback, (c) reinforcement, and (d) dosage can be faded (Table 1).
Maintenance Programming Examples for Tier 2 Interventions
James’s PBIS team decides to implement a multicomponent self-monitoring intervention during math, the class in which he struggles the most. For 4 weeks, James has been self-monitoring his behavior every 5 minutes the entire 45-minute math period. James thinks about and records whether or not he displayed the PBIS schoolwide expectations of “Be Responsible” and “Be Respectful” during each 5-minute interval. In addition to James monitoring his own behavior, his math teacher (Ms. Hoya) has been monitoring James’s behavior using the same rating and interval system as James. After every math class, Ms. Hoya gives James behavior specific feedback about his performance. Additionally, if James meets his performance goal for the day, he is given a PBIS ticket, which is part of the schoolwide reinforcement system.
Goals
Students often set behavioral goals related to their performance in CICO, small-group social-skills instruction, or self-regulation interventions. For instance, students might set a goal to receive a certain number of points on the CICO report card. Students receiving social-skills instruction may set a goal to increase the number of positive interactions they have with their peers. Or students self-monitoring their on-task behavior might aim to be on task for a certain number of intervals or percentage of time. It is imperative for teachers to frequently track students’ progress toward the goal to make data-informed decisions. When students are responding well to interventions consistently (e.g., scores falling within 5% of the goal for 3 to 5 consecutive days), teachers may consider talking to the student about raising the goal to a mutually agreeable level that is no more than 10 percentage points above average performance from the previous 5 days (Bruhn et al., 2017). One of the hallmarks of goal-setting interventions is that goals are attainable. If a goal is raised too quickly or too high and becomes unattainable, it is likely the student will become frustrated and may quit trying to achieve the goal (Locke & Latham, 2002). Research has demonstrated raising goals incrementally over time in CICO and self-monitoring interventions results in students’ achievement of their goals and maintenance of high rates of positive behavior (Bruhn, Vogelgesang, Fernando, & Lugo, 2016; McDaniel & Bruhn, 2015).
Since the launch of intervention, Ms. Hoya has been reviewing James’s behavior every 5 days. Prior to intervention implementation, James’s initial behavior performance was 40%, calculated as a daily, average score of the two behaviors being monitored (i.e., “Be Responsible” and “Be Respectful”), collectively called positive behavior. James has been responding well to the intervention, so every 5 days his goal has been increased by 10 percentage points from the initial goal set at 50%. Currently, his goal is set at 70%, and he is scoring between 70% and 73% on positive behavior. Ms. Hoya is thrilled with this improvement in behavior, but she would like to see him performing around 80% to 90%, a rate similar to that of other peers in his class. Ms. Hoya decides to increase James’s goal to 80%.
Ms. Hoya determines it would be best for both James and for the classroom environment to have him perform at the same level as his peers. By reviewing the data and seeing James’s scores fall between 70% and 73% on positive behavior for 5 consecutive days, Ms. Hoya sees James’s behavior change has been consistent and the goal can be increased. Additionally, Ms. Hoya decides to increase the goal by only 10 percentage points above the average of James’s prior 5 days, making the new goal realistic and attainable.
Ms. Hoya meets with James to talk about the newly set goal. Ms. Hoya shows James his data and tells him she is really happy with his progress. She then asks James if he thinks he can increase his scores to around 80%; James agrees that this is a reasonable increase.
When goals are increased, teachers may (a) increase the goal independently, (b) include student input, or (c) have students self-select their goal. Although any method may work, research suggests when students have direct input on their goals, they are more committed to achieving the goal (Bruhn et al., 2017). If a teacher decides to have students self-select their goal, the teacher should provide guidance by showing students their data to help them make an informed decision that is appropriate and attainable.
As progress is made toward improving desirable behaviors, knowing when to stop increasing the goal depends upon the behavior. In many cases, setting goals to 100% (for positive behaviors) or 0% (for negative behaviors) is inappropriate, as perfection is not a reasonable expectation. Goals should continue to be increased until the student has reached the desirable level of behavior for a sustained period of time. For Ms. Hoya, it was important for her to have James performing at similar levels to peers in the class; therefore, she stopped increasing the goal once James was consistently scoring 80%.
Feedback
Throughout interventions, teachers may decide to provide behavior-specific feedback to students. Feedback can be in the form of behavior-specific praise or error correction. In other words, a teacher may tell the students what exactly they did well (e.g., “Great job working on your math sheet the whole time!”). Or a teacher may need to calmly and kindly tell students what they did wrong and what they should do instead (e.g., “I see that you left your chair in the aisle after you got up. Please push your chair in like the rest of your table team.”). Regardless of the type of feedback, it is important the feedback is behavior specific. Conversely, when a teacher tells a student, “Good job,” the statement does not specifically inform the student of the good behavior the teacher noticed, nor does the student know what behavior to continue doing. Rather, the teacher should say, “Maria, good job raising your hand to ask a question.” In this statement, the student knows the behavior for raising her hand was a behavior she was doing well. Not only does this increase the chances that Maria will raise her hand again in the future, but other students may overhear this praise and then display appropriate behavior.
Feedback can serve several purposes, such as increasing the likelihood that the student will behave positively while also providing students with attention. Some students may thrive on teacher attention, especially if attention is the function of their behavior. Fading feedback simply involves reducing the dosage of feedback. For example, feedback is a core component of CICO. Teachers initially provide feedback every time they complete the daily report card, which occurs multiple times throughout the day. If the student is responding well, then teachers could reduce the number of times they provide feedback when completing the daily report card. Or they could decrease the number of times the daily report card is completed, thereby reducing feedback. This fading strategy has resulted in sustained reductions in problem behavior and increases in academic engagement that maintained as feedback was faded (Campbell & Anderson, 2011). In social-skills instruction, best practice involves providing high rates of feedback when students are practicing or role-playing as well as when those skills are displayed outside of small-group instruction (e.g., on the playground, during cooperative learning activities). As students demonstrate greater fluency with specific skills, teachers can gradually reduce the amount of feedback they provide. Throughout the fading process, it is essential to continue monitoring student behavior to ensure positive behavior is maintaining. Further, feedback (i.e., specific praise and error correction) is considered an evidence-based Tier 1 support, so teachers should avoid fading feedback completely (Perle, 2016). Instead, feedback should be continued intermittently, as needed.
James and Ms. Hoya meet after every math class, where Ms. Hoya provides James with behavior-specific feedback. Ms. Hoya is sure to include both praise and corrective feedback statements as appropriate. These meetings are quick, about 1 minute in length. They provide time for Ms. Hoya and James to build a relationship and provide specifics on how James can improve and maintain his behavior. James seems to enjoy the teacher attention, yet Ms. Hoya knows she has to fade this support so that James isn’t totally reliant on her. Instead of meeting after every session, Ms. Hoya decides to meet every other day to provide feedback. Assuming James maintains his behavior, Ms. Hoya will fade feedback meetings to once a week and eventually to no meetings at all. At that point, she can deliver feedback as needed.
Reinforcement
Interventions commonly include reinforcement, or rewards, given to students for meeting some criterion. This means the student must perform at a certain level or for a specified amount of time to receive reinforcement. For instance, a teacher could provide a ticket every time the target student raises a hand to ask a question, thus reinforcing hand-raising behavior. Or a ticket may be provided to students who remain in their desks for the duration of independent work time. Although a wide range of reinforcers exists (e.g., tangible, sensory, activity, attention, escape), it is important for students to have input and choices about potential reinforcers. If a teacher selects something that the student is not motivated to earn, then the reinforcer is unlikely to have a positive effect on behavior.
Like feedback, fading reinforcements from an intervention involves reducing the frequency of reinforcement by changing the contingencies. For example, if a student is using CICO and receives a PBIS ticket every day for meeting a target goal and is highly successful, the teacher might choose to provide PBIS tickets for meeting the goal for 2 consecutive days. If the student continues to reach the goal, then the student might earn a PBIS ticket for meeting the goal 4 out of 5 days for the week. Research has shown changing the reinforcement contingency in CICO once students are responsive can help to maintain positive behaviors (Miller, Dufrene, Sterline, Olmi, & Bachmayer, 2015). The idea is to completely remove the need for the student to receive something (e.g., break, ticket, toy) for displaying the appropriate behavior and move to more natural forms of reinforcement, such as specific praise (Cooper et al., 2007). As with every other component, student data should be collected and evaluated continually to monitor the student’s response to intervention.
When James meets his goal in Ms. Hoya’ math class, he earns a 5-minute computer break at the end of class. To fade this component, Ms. Hoya increases the contingency required to receive the break. Now, James has to meet his goal 2 days in a row to receive his break.
Dosage
The dosage of the intervention refers to the amount of time or how often intervention components are implemented. As students respond to the intervention, the dosage should be reduced. In CICO, for example, instead of giving the daily point sheet (or report card) to every teacher throughout the day, a student could give the point sheet to the teachers in morning classes only—the classes where problematic behavior was most likely prior to the start of the intervention. For interventions that require small-group instruction, such as social skills, instead of 45-minute instructional sessions, they could be reduced to 30-minute sessions. Or instead of meeting daily, groups could meet once or twice a week, which has shown to be an effective way to sustain appropriate social-skill use (Peterson, Young, West, & Hill, 2006). Similarly, self-monitoring interventions can be adapted by gradually increasing the length of intervals, thereby reducing the number of opportunities to self-monitor (Rock & Thead, 2007).
Ongoing data collection is foundational to informing intervention decisions…consistent procedures for collecting and analyzing data must be in place.
James is self-monitoring every day during math on a 5-minute interval schedule. James has been responding well to the intervention over the weeks, and Ms. Hoya decides to increase the interval length to 10 minutes. Ms. Hoya continues to monitor James’s behavior, which remains steady for the next 5 days. It is clear to Ms. Hoya that James’s intervention needs to be reduced further. She then decides to have James monitor for 3 days a week instead of 5 days, still using the 10-minute intervals. Ms. Hoya will continue to monitor James’ behavior until the self-monitoring intervention can be completely removed.
Other Intervention Components
Other intervention components might include the inclusion of role-playing during social-skills instruction, parental signatures on CICO report cards, or self-graphing progress toward goals. Like the aforementioned components, the frequency and dosage of these components need to be slowly reduced over time. This means, perhaps, role-playing is removed from social-skills lessons. Or as explicit social skills instruction is reduced, the teacher uses incidental teaching (i.e., “teachable moments”) to informally address social skills as they naturally occur (Gresham, Horner, & Sugai, 2002). Parent signatures may be required weekly instead of daily in CICO, or self-graphing may be eliminated from self-regulation interventions. Teachers can determine which components to systematically fade based on student data and student characteristics. Perhaps a student does not like role-playing or is inconsistent with having parents sign the CICO daily report card, so it makes sense to fade these components first. Of course, teachers need to ensure students’ positive behavior maintains. Without monitoring student progress, the teacher cannot determine the extent to which the student is responding to the newly modified, less-intense intervention.
Suppose James’s positive behavior continued to increase while self-monitoring only 3 days a week. Ms. Johnson could choose to cease self-monitoring all together but still provide James his weekly feedback meetings. Ms. Johnson feels that this modification is sustainable and appropriate for long-term use, therefore returning James to only Tier 1 supports.
Final Thoughts
There is not an exact recipe or formula for programing an intervention for sustained behavioral improvement, especially considering (a) the wide variety of components that comprise Tier 2 interventions and (b) the diversity of individual student needs. Whereas a teacher can focus on reducing or removing one component at a time, the teacher also may decide to fade various intervention components in a staggered manner, meaning a teacher can first fade one component (e.g., changing the reinforcement contingency) and then, if the student is responsive, decide to fade a different component (e.g., reduce the dosage). The teacher can toggle between the two different components or continue with one component and then move to fading another component until they are completely eliminated.
Given the flexibility of adapting interventions and the need to respond according to the data, this process is unlikely to proceed linearly. That is, ongoing data collection is foundational to informing intervention decisions. As part of implementing and evaluating any intervention, consistent procedures for collecting and analyzing data must be in place (Bruhn et al., 2018). Generally, examining behavioral data (e.g., direct observation data, direct behavior ratings) every 3 to 5 days can help teachers determine how a student is progressing (or not) with intervention (Bruhn et al., 2018). Teachers can examine whether students have met behavioral goals, if behavior is becoming more stable, if there is an overall change in the level of behavior, or if behavior is trending in the right direction. When student response is positive and consistent for a sufficient period of time (e.g., 2 to 3 weeks) with the intervention in place, then a plan for slowly reducing or removing intervention can be put in place (see Figure 1).
Systematic fading for responders to intervention
In sum, as students improve their behavior through Tier 2 interventions and supports, it is important that teachers systematically fade interventions to encourage long-term behavioral change. Although this process may seem difficult because it is nonlinear, has no hard and fast rules, and involves some experimentation, positive behavior is more likely to maintain with systematic fading than when a practitioner simply stops an intervention “cold turkey” after the student has shown improvement. Through systematically reducing and removing intervention components, such as (a) goals, (b) feedback, (c) reinforcements, (d) dosage, or (e) other intervention elements, teachers can increase the likelihood behavioral change will maintain as students transition to only Tier 1 supports.