Classroom-based functional analysis and intervention for disruptive and off-task behaviors.Abstract
Although there is a growing body of literature on the use of functional analysis in schools, there is a need for more demonstrations of this technology being used during the course of typical instruction. In this study, we conducted functional analyses of disruptive and off-task behavior in a reading classroom setting for 3 participants of typical intelligence identified as at-risk for reading failure. The teacher implemented two functional analysis conditions (i.e., escape, and attention) and a control condition. The results of the functional analysis suggested that disruptive behaviors were maintained by teacher attention for all 3 participants. Based on the functional analysis results, the teacher implemented a differential reinforcement of other behavior procedure in which participants were given a high rate of attention in the absence of disruptive and off-task behaviors. In addition, the teacher implemented a differential reinforcement of alternative behaviors procedure in which the appropriate recruitment of teacher attention was reinforced. The procedures were implemented using a multiple baseline design across participants. During intervention, the disruptive behaviors of all 3 participants decreased to near zero levels.
The need to conduct functional behavioral assessments (FBA) in general education settings is increasing as the number of behaviorally diverse students increases in schools. This increase is due partly to the mainstreaming and full inclusion of children with developmental disabilities and severe behavior problems (Scott et al., 2004). Furthermore, the majority of school-based discipline referrals are from children engaging in distractive or disruptive behaviors (Sterling-Turner, Robinson, & Wilczynski, 2001). Federal law mandates that children engaging in challenging behaviors have supports put in place to keep them in their current placement and to enable them to make sufficient academic progress (No Child Left Behind Act of 2001). The mandated supports include the use of FBA and positive behavior support plans to address the challenging behaviors presented by students in school settings (Individuals with Disabilities Education Act ([IDEA], 1997/2004). IDEA does not specify exactly what represents a valid FBA; therefore there is currently no single protocol for schools to follow. The lack of a protocol has led to some confusion about what actually constitutes a FBA and what the differences are between it and an experimental functional analysis (EFA; Gresham, Watson, & Skinner, 2001).
A FBA typically describes a range of indirect and direct procedures (e.g., interviews, questionnaires, descriptive analyses, direct behavioral observations, experimental functional analyses) that can be used to identify potential antecedents and consequences associated with the occurrence of problem behaviors (Gresham, et al., 2001). An EFA refers to the experimental manipulation of antecedent and consequent events in a controlled setting to identify functional relations between environment and behavior (Iwata et al., 2000). Horner (1994) suggested that an EFA is but one approach to FBA. However, an EFA is the only approach to FBA that uses experimental manipulations to find an empirically supported function of aberrant behavior rather than descriptive or correlational hypotheses about the operant function of behavior (Carr, 1994). The basic EFA methodology developed by Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994) has been applied in various forms (Broussard & Northup, 1997; Mace & West, 1986; Mueller, Sterling-Turner, & Moore, 2005), but the basic methodology remains the same: by controlling the environment using the EFA methodology (e.g., the rate of reinforcement delivery, control of extraneous variables), the maintaining variables of the problem behavior can be empirically identified.
In addition to the need to identify the function of a student's problem behavior, there is also a need to implement function-based interventions for children in school and classroom environments (Scott & Kamps, 2007). Gresham and colleagues (2001) suggested that many general classroom interventions may be ineffective because they are not based on the function of the students' behavior. When interventions are not based on the function of the problem behavior, they typically rely on strong reinforcers and/or punishers that may override the current contingencies maintaining the problem behavior (Mace, 1994). Vollmer and Northup (1996) identified several problems that can arise when interventions are selected without considering behavioral function. These problems include strengthening the problem behavior via positive or negative reinforcement. In addition, the intervention may not be functionally related to the contingencies maintaining the problem behavior and the intervention may not address alternative behaviors that are more socially appropriate that serve the same function.
Moreover, problem behaviors that appear to be similar may occur for significantly different reasons. Vollmer and Northup (1996) summarized that there are three common sources of reinforcement of problem behaviors in the classroom: attention from the teacher, attention from peers, and escape from instructional demands. These contingencies are readily available in the classroom and therefore available to be used in the context of an EFA.
An emerging theme in the current EFA literature is for the student's educational personnel to conduct the EFA and implement the function-based intervention (Bessette & Wills, 2007; Ervin et al., 2001; Kamps, Wendland, & Culpepper, 2006; Sasso et al., 1992; Scott et al., 2004; Wright-Gallo, Higbee, Reagon, & Davey, 2006). While there is still a limited amount of research on school personnel in educational settings conducting EFAs (Hoff, Ervin, & Friman, 2005), the body of literature is quickly growing (Bessette & Wills, 2007; Erbas, Tekin-If-tar, & Yucesoy, 2006; Kamps et al., 2006; Wright-Gallo et al., 2006).
Another area of limited research is the use of EFA methodology with typically developing children, and more specifically, with children who have not been identified with any type of disability, but who are displaying aberrant behaviors (Moore, Edwards, Wilczynski, & Olmi, 2001). Most of the current literature focuses on using EFA methodology with children with a wide range of disabilities, especially children with developmental disabilities (Hanley, Iwata, & McCord, 2003). Moore and Edwards (2003) evaluated the influence of social variables and task-difficulty in relation to escape maintained problem behavior with 4 students in their typical classroom (only one was identified as having a diagnosis, severely emotionally disturbed, and the other three were identified as typically developing). They found that for 2 students, the cessation of teacher reprimands was believed to have served as a negative reinforcer, resulting in a decreased rate of disengagement (respectively increasing engagement) and a decreased rate of problem behaviors. For the other 2 students, praise (e.g., encouraging verbal statements) provided by the teacher contingent on student engagement, was believed to have served as a positive rein-forcer, resulting in a higher rate of engagement (respectively decreasing disengagement) and a lower rate of problem behaviors. This study demonstrates the usefulness of employing EFA methodology with typically developing children. More explicitly, the results from the EFA made it possible for the authors to discriminate between types of teacher attention (i.e., in the form of reprimands and praise) serving as negative reinforcement for disengagement and problem behaviors and teacher praise serving as positive reinforcement for engagement and desirable behaviors.
The purpose of the current study was to evaluate the implementation of an EFA and function-based intervention by the regular school personnel with 3 typically developing children. Data were collected and used to answer the following questions:
1. To what extent can a brief EFA, run in the general education classroom by school personnel, be effective at identifying the function of disruptive and off-task behaviors during reading time for 3 typically developing participants?
2. To what extent can the intervention based on the results of the EFA be effective at reducing disruptive and off-task behaviors and increase the engagement of the 3 participants during academic tasks?
3. To what extent does the teacher rate the EFA and intervention procedures as socially acceptable?
The study took place in an urban elementary school with 278 students. A total of 65% of students received free or reduced lunches. All 3 participants were in the same second grade reading instruction rotation class with six other children, the teacher, and a paraprofessional. The reading block for second graders was split into one 45 min large group general reading instruction period and two 20 min periods. During one of the 20 min periods all students participated in a computer-based reading program and during the other 20 min period, all 3 participants participated in a rotation of instruction using the Programmed Reading curriculum (Sullivan & Buchanan, 1988). Programmed Reading is an individualized reading curriculum designed to be used independently by students. Each student was assigned a book on his or her individualized instructional reading level. Students were instructed to read quietly aloud, while the teacher monitored each student for approximately 2 min per day. The teacher in this setting reported that 3 students were exhibiting high rates of disruption and were frequently off-task. All experimental sessions (i.e., baseline, EFA and intervention) occurred in the regular reading room during this 20 min Programmed Reading rotation.
Two males and one female identified as at-risk for later reading failure based on scores on the Dynamic Indicators of Basic Early Literacy Skills (DIBELS; Good & Kaminski, 2002), Brandon, Paul, and LaTonya, served as participants in this study. Brandon was a 7-year-old male, Paul was an 8-year-old male, and LaTonya was a 7-year-old female. The teacher reported that all 3 participants engaged in high rates of disruptive behavior in addition to high rates of off-task behavior during the 20 min daily small group Programmed Reading sessions. Examples of the disruptive behaviors that influenced the teacher's decision to refer these students included calling out to the teacher from across the room, out-of-seat behavior, and general off-task behavior (e.g., "staring into space"). None of the participants were receiving special education services; they were all in a regular education classroom without additional supports. Furthermore, none of the participants were reported to have been diagnosed with any psychiatric conditions according to the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders (4th ed. TR, 2000).
DIBELS testing was conducted three times a year at the participants' school. Brandon was absent during the beginning second grade DIBELS testing. His score for oral reading fluency during April of first grade was 30 correct words per minute. The goal is 40 or more correct words per minute for oral reading fluency at the end of first grade. Paul and LaTonya were tested in September of second grade. Paul's score for oral reading fluency was 23 correct words per minute. LaTonya's score for oral reading fluency was 26 correct words per minute. The goal is 42 or more correct words per minute for oral reading fluency at the beginning of second grade.
The teacher was a 27-year-old male whose official role in the school was as a reading intervention provider. The teacher had been at the current school for 3 years implementing small group reading interventions and although employed as a reading teacher he was still completing his certification requirements. There were a total of 9 students, 3 females and 6 males in the classroom with the teacher and paraprofessional. The teacher and the paraprofessional rotated around the group of 9 students who were spread out among the desks available. Before the current study began, the paraprofessional was instructed to not engage with the 3 study participants for the remainder of the school year and to only rotate among the other 6 students. This was to help control for the other adult present in the room who was not participating in the study so that attention or demands from the paraprofessional did not confound the results.
Student behaviors. The primary dependent variables were disruptive and off-task behaviors. Disruptive behaviors included behaviors that appeared to interfere with learning, impede instructional delivery, or both. Disruptive behavior included the student arguing, taunting, name calling, making audible vocalizations unrelated to the instructional task (i.e., singing, humming, and talking to self), making repeated audible noises with tangible items (e.g., pencil tapping), talking to peers, calling out the teacher's name with or without hand-raising, getting out of their seat and walking up to the teacher during seat work, and waving their hand in the air. Non-examples included responding appropriately to the teacher's verbal questions, talking with teacher permission, and quietly reading out loud. Off-task behavior was defined as not attending to or participating in instructional activities as requested by the teacher. Examples of off-task behaviors included engaging in disruptive behaviors (as defined above) while not engaging in the academic task. These definitions were not mutually exclusive. For example, if the participant was engaging in pencil tapping, but was engaged in the academic task, then only disruptive behavior was scored. Other examples of off-task behaviors include gazing around the classroom and not following instructional directions. Non-examples included doing or attempting the assigned task (e.g., quietly reading out loud, looking at the teacher or instructional activity), seeking assistance appropriately, and following directions.
Teacher's behavior. In addition to the measurement of the student's behavior, data were collected on teacher attention to students and demands to students. Examples of teacher attention included any verbal comment directed to the participant (e.g., praise, answering questions, correcting, or reprimanding), or gestures (e.g., high fives, pats on the back, eye contact with a nod of the head, and a thumbs up). Demands were defined as statements to engage in an academic activity. Examples include, but were not exclusive to statements such as "begin reading," "open your books and read," and "read out loud." The purpose of collecting data on the teacher providing attention to appropriate and inappropriate behaviors and giving demands was to monitor the integrity of implementation during the experimental analysis and intervention conditions.
All experimental sessions were 5 min in duration. The researcher used a momentary time sampling procedure to collect off-task data, with data being collected at the end of 10 s intervals. The researcher used a 10 s partial interval recording procedure to collect all other behavioral data. During baseline and intervention, only one 5 min observation was conducted with each child who was present each day, with the exception of instances when there was insufficient time to run a full observation with each participant.
The reading teacher implemented both the EFA conditions (under the guidance of the experimenter), and the subsequent function-based treatment during the intervention condition. All baseline, EFA conditions, and intervention sessions occurred while the participants were in the reading room during the Programmed Reading instructional time period. Participants were seated at their normal desks during all sessions except for the EFA control/play condition during which they moved to the reading corner. Participants were generally assigned to independent reading, with some variations in terms of activities, such as the participant reaching a test page and earning a free day. When the participants earned a free day they were allowed to spend the following day in the reading corner reading books of their choice for the entire 20 min period.
For the purposes of monitoring the teacher's integrity of implementation, the researcher and secondary observers recorded if the teacher implemented the correct contingency related to the participant's appropriate or inappropriate behavior during the EFA and intervention conditions. Each 10 s interval was scored as correct or incorrect and the criteria varied based on which contingencies the teacher was instructed to implement for each condition. If the teacher responded correctly and incorrectly within the same interval, the interval was scored as incorrect.
The experimental design was a multiple baseline design across participants. The sequence of procedures during the multiple baseline consisted of collection of baseline data for all participants, then the teacher implemented the functional analysis for the first participant, followed by implementation of the intervention for the first participant. Baseline data were also collected for the other 2 participants on the same day that intervention data were collected for the first participant. Next, the teacher implemented the functional analysis for the second participant, followed by intervention for the second participant and so forth for the third participant. The order of observations was randomized across participants and the teacher was unaware of which participant was being observed at any given time as data were being collected for all participants, each day in their respective baseline and intervention phases. The only exception was during the functional analysis when data were only collected for 1 participant across all three conditions. Therefore, the teacher was aware on whom data were being collected during the functional analysis sessions.
Training secondary observers to criterion levels of 80% agreement achieved inter-observer agreement. Observers reviewed operational definitions of student and teacher behaviors and then collected data in the classroom until the criterion level was attained for all target behaviors.
Agreement was defined as both observers scoring either an occurrence or nonoccurrence for each target behavior during each 10 s interval. Inter-observer agreement was collected using a point-by-point agreement method. The number of agreements was divided by the number of agreements plus disagreements. These quotients were then multiplied by 100.
A second observer collected data during 25% of observations across all conditions (i.e., baseline, functional analysis, and intervention) and participants. Inter-observer agreement remained consistently high across all conditions and did not vary by condition. An average agreement score was calculated across all conditions and all 3 participants, mean = 98%. For Brandon, inter-observer agreement data were collected during 25% of observations across all conditions (mean = 99%, range = 99%-100 %). For LaTonya, inter-observer agreement data were collected during a total of 27% of observations across all conditions (mean = 98%, range = 97%-99%). For Paul, inter-observer agreement data were collected during a total of 24% of the functional analysis and intervention observations (mean = 98%, range 95-100%). Due to Paul's absence from school on the days that a second researcher was present in the classroom, no inter-observer agreement data were collected during his baseline observations. Mean inter-observer agreement for teacher behaviors included teacher attention to appropriate behaviors (mean = 99%, range = 98%-100%), teacher attention to inappropriate behaviors (mean = 99%, range = 99%-100%), and teacher presentation of demands (mean = 99%, range = 98%-100%).
Procedures: Functional Behavioral Assessment
Functional assessment interview. Prior to the descriptive assessment, the researcher interviewed the reading teacher using a modified version of the Functional Assessment Interview (FAI; O'Neill et al., 1997). Modifications included the exclusion of questions concerning the person's sleep schedule, medications, and eating habits. This interview was conducted to gather information regarding each student's problem behaviors and possible contingencies maintaining those behaviors. During the interview, which lasted approximately 30 min, the teacher and researcher formulated hypotheses regarding the function of the participant's disruptive and off-task behaviors. They also discussed and defined the disruptive and off-task behaviors that would be targeted during the study.
Direct behavioral observations (descriptive baseline). Prior to the experimental functional analysis, a descriptive assessment was conducted using direct behavioral observations of the reading teacher and each participant in the reading classroom. All observations were 5 min in length. All observations were conducted in the same classroom in which the functional analysis and intervention were later conducted. The researcher was seated on the perimeter of the classroom during all sessions. The rationale for the descriptive assessment condition was to document pre-functional analysis levels of all dependent and independent variables so that comparisons could be made to post intervention levels of these variables. The descriptive data provided information regarding the rates of the participant's problem behaviors and antecedents prior to and consequences following each occurrence of problem behavior. These data aided in the development and implementation of the experimental functional analysis conditions by providing data in support or opposition of the hypotheses generated during the teacher interview. It also helped to address the concerns of school personnel regarding the direct manipulation of antecedents and consequences that could possibly increase problem behaviors. These data are reported as the baseline rate of disruptive and off-task behavior.
Procedures: Teacher Functional Analysis Training.
Following the functional assessment interview and observations for the first participant, the researcher taught the teacher how to conduct a functional analysis for three conditions. In addition to a control play condition, the only maintaining conditions for off-task and disruptive behavior indicated by both the teacher interview and the observations were attention and escape. During the 10 min training, the researcher provided the teacher a brief descriptive statement for each condition printed on three different colors of 8.5 x 11 inch paper. On the yellow sheet of paper the following directions were printed: "Attention Condition: Attend to all inappropriate behaviors, Ignore all appropriate behaviors." On the blue sheet of paper the following directions were printed: "Escape Condition: Every 30 seconds give a clear instruction 'Name, start reading.' If the student complies within 5 seconds praise, if noncompliant repeat instruction, if still noncompliant, take away materials and ignore until the next instruction. If the student engages in a disruptive behavior, remove the task until the next instruction." On the red sheet of paper the following directions were printed: "Play Condition: No demands (except for the initial instruction), attention given every 30 seconds, if an inappropriate behavior occurs do not give attention until 5 seconds after the behavior. When the student tries to interact appropriately, reciprocate. Ignore all inappropriate behaviors."
During the 10 min training, the researcher described different likely scenarios and role-played the correct responses. The researcher modeled examples of each condition and involved the teacher in practice with feedback. The teacher's questions and concerns were addressed as they arose during the training. Finally, the definitions of behaviors were again reviewed. Additionally, prior to and following each session, the teacher was asked if he had any concerns or questions and these were addressed as they arose.
Procedures: Functional Analysis
Experimental conditions were conducted with all 3 participants (escape, control/play, and attention). These conditions were based on those developed by Iwata et al., (1982/1994) (see description of conditions below) with some notable exceptions. First, all conditions were conducted during the participants' normal academic time in a small group, individually paced reading rotation. Second, unlike Iwata et al.'s control/play condition during which no demands were given, at the beginning of the control/play condition, participants were instructed to pick from a variety of books and to continue to read while in the reading corner. Additionally, in the current study a demand was always placed at the beginning of each class to read because it was the participants' normal academic instruction time; the reading materials were present during the attention condition; and neither a tangible nor an alone condition were conducted with any of the participants. These conditions were not included because the initial hypotheses based on both the teacher interviews and student observations prior to the study indicated that the likely maintaining function for all participants was either attention or task escape.
Attention. During this condition, the teacher engaged in his usual teaching practice, which involved rotating around the room listening to students read while monitoring other students in the group simultaneously. The paraprofessional was also rotating around the room, but remained near the other students in the classroom and did not stand near the participants of this study. Therefore, the teacher stayed close to where the participants were seated and was always the one to respond when the participants raised their hands. With the paraprofessional and teacher placement, it is unknown if these variables possibly influenced the participants' behavior, although this change to the environment was consistent across all phases of the study.
At the beginning of each rotation, the teacher instructed the class to begin reading. Beside the initial demand to read, the teacher was instructed not to place demands during this condition. After giving the demand to start reading, the teacher monitored the classroom. Each time a participant engaged in a disruption or was off-task the teacher gave immediate brief attention. Verbal attention consisted of brief statements in response to the participant. For example, when a participant was off-task the teacher would ask the participant if they were having trouble or why they were not reading. When a participant engaged in calling out for the teacher, the teacher would immediately respond by asking them what they needed. If the participant walked up and asked the teacher a question, he would briefly respond to the participant's question and then continue with what he was previously doing. If the participant again asked the teacher a question, the teacher would repeat the above protocol. In addition to providing verbal attention, the teacher also temporarily moved closer to the participant. There was not a set protocol for how close the teacher needed to move, this was only done to minimize the amount of disruption to the other students in the classroom. During the attention condition, the participants' reading materials remained on the participants' desks the entire session and were not removed contingent on disruptive or off-task behavior.
Escape. During this condition, the teacher prompted the participant to engage in reading every 30 s. If the participant began/continued to read, the teacher provided a brief verbal praise statement. If the participant did not begin reading within 5 s, the teacher provided another prompt. If the participant began reading after the second prompt, the teacher provided a brief verbal praise statement (e.g., "good job reading"). If the participant did not begin reading within 5 s after the second prompt, the teacher removed the reading materials from the participant's desk and returned the materials at the beginning of the next planned 30 s prompt. In addition, the teacher turned away from the participant and ignored all behavior until the next trial (e.g., no attention was provided during the escape interval). The escape condition differed from the attention condition in two aspects. First, during the escape condition, teacher attention was delivered in the form of the demand to read, then in the form of verbal praise contingent on the participants reading within 5 s of the demand being placed otherwise if they engaged in targeted behaviors the reading materials were removed and no further teacher attention was provided. Therefore, unlike the attention condition when attention was contingent on problem behaviors, attention was only provided in the form of the demand and in the form of praise for on-task behavior and escape was contingent on the targeted behaviors. Second, while the demand to read was stated at the beginning of each class, it was only repeated every 30 s during the demand condition and not the other conditions.
Control/Play. During this condition, the participant was informed that for the next few minutes they would be allowed to engage with a book of their preference in the reading corner. This was a small area in the front of the classroom that had a bookcase, books, blanket and pillows. During normal teaching activities, all students spent the 20 min reading period in the corner reading books they had chosen from the bookcase the day after passing a test. While the participants were in the reading corner during the EFA conditions, the teacher provided attention (a brief verbal praise statement) delivered on a 30 s schedule. If the participant recruited the teacher's attention appropriately, the teacher responded immediately. A 5 s time delay for teacher attention was implemented if the participant engaged in any inappropriate target behaviors. The delay was implemented during the control/play condition to control for target behaviors inadvertently being reinforced by teacher attention.
The conditions during each participant's EFA were alternated randomly, with the exception of not conducting the conditions in the same order for two sessions in a row. This was to help control for order effects. Each of the three conditions was conducted once per day until clear patterns were observable through visual inspection of the graphed data. Based on the patterns observed for all 3 participants, only three sessions with all three conditions were conducted for each participant. For each participant the EFA took a total of 45 min, 15 min each day for a total of 3 days.
During the experimental conditions, the teacher held the corresponding color coded instruction sheet that had the instructions for the condition which he was conducting to enhance discrimination between different functional analysis conditions and provide a visual prompt for the teacher. To decrease the disruptions of the researcher prompting the teacher during the functional analysis, the researcher held the corresponding color-coded instruction sheet. The researcher signaled that the session had ended by moving the corresponding color-coded sheet from left to right. Following the signal, the change of color-coded sheet served as another visual signal to the teacher which condition to conduct next. Prompts during sessions included signals as needed to provide attention every 30 s during the play condition, to return the task during the demand condition, and to present demands.
Based on the results of the functional analyses for all 3 participants, teacher attention was found to be the maintaining variable for disruptive and off-task behaviors during the reading rotation. The intervention for all 3 participants was designed to target the maintaining variable of teacher attention. The intervention consisted of two behavioral components. The two components were differential reinforcement of other behaviors (DRO) with extinction, in conjunction with differential reinforcement of alternative behaviors (DRA). With the DRO component, the teacher used a self-monitoring form to monitor his delivery of attention approximately once per min as long as the participants had not engaged in inappropriate target behaviors within 5 s prior to the end of the 1 min interval. The teacher marked on the self-monitoring form each time he provided attention to each participant. This was a permanent product that the teacher could refer to during the reading rotation to monitor his rate of praise and adjust if necessary. Because both target behaviors (e.g., disruptive and off-task behaviors) were maintained by the same variable (teacher attention), it is believed they were both in the same response class. Therefore, the teacher was instructed to not provide attention within 5 s of a participant engaging in any targeted behaviors (disruptive or off-task). For the DRA component, the teacher immediately responded to the participants' raising their hand without engaging in disruptive behavior. The teacher responded by praising the participant and providing assistance with their request. The effectiveness of the DRO and DRA intervention was evaluated using a multiple baseline across participants design.
Follow-up sessions were not possible due to the participants extended absences during intervention, in addition to the end of school quickly approaching and the high rate of field trips and assemblies at the end of the school year.
Social Validity Questionnaire
The teacher completed a satisfaction survey developed by the researchers at the end of the study for each participant, using a 5-point Likert scale. The survey consisted of nine statements to be rated on a scale of one to five (1 = strongly agree to 5 = strongly disagree). In addition, there were four-open-ended questions for the teacher to complete and another space requesting other comments and suggestions.
Functional Behavioral Assessment
Results from the interviews suggested problem behaviors were disruptive behaviors and off-task behavior for all 3 participants. When asked to identify the "function" of each participant's disruptive and off-task behaviors, (i.e., when asked, "After the occurrence of a problem behavior, what does the participant gain or avoid?") the teacher responded that for Brandon and Paul it was to avoid work. The teacher reported that for LaTonya the function of her disruptive and off-task behavior was to gain attention and to avoid work. This was inconsistent with the researcher's hypotheses based on the direct behavioral observations in which attention appeared to be the primary maintaining function, but the researcher could not rule out an escape function. Therefore, a functional analysis was needed to empirically identify the function of disruptive and off-task behavior. It should be noted that both a tangible and automatic function were not found to be possible functions based on the teacher interview and direct behavioral observations.
Experimental Conditions: Baseline, EFA, and Intervention
Figures 1 and 2 represent the naturalistic baseline, EFA, and intervention data for each participant. All data are reported in percentages to allow for analyses. Figure 1 represents the percentage of intervals with disruptive behaviors and Figure 2 represents the percentage of intervals with off-task behaviors. For Brandon, as depicted in Figure 1, there was an upward trend for disruptive behaviors during baseline. During the EFA, there was an elevated percentage of intervals with disruptive behaviors in the teacher attention condition while maintaining zero to near zero percentages during the escape and play conditions. These results suggest that attention was the maintaining variable for Brandon's problem behaviors. As depicted in Figure 2, there was an upward trend during baseline and a high level of off-task behaviors during the attention condition, while maintaining a low level of off-task behavior during the escape and play conditions, further supporting that attention was the maintaining variable for Brandon's problem behaviors. After the DRO + DRA intervention was implemented, disruptive behaviors decreased and maintained at zero to near zero levels, while off-task behavior decreased to near zero and maintained near zero levels.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
For LaTonya's behavior, as depicted in Figure 1, there was an upward trend for disruptive behaviors during baseline and during the EFA there was an elevated percentage of disruptive behaviors in the teacher attention condition while maintaining a low percentage of intervals during the escape and play conditions, suggesting that attention was the maintaining variable for LaTonya's problem behaviors. In Figure 2, there was an upward trend during baseline and a high and variable percentage of intervals of off-task behaviors during the attention condition, while maintaining a low level of off-task behavior during the escape and play conditions, further supporting that attention as the maintaining variable. After the DRO + DRA intervention was implemented, disruptive and off-task behaviors decreased and maintained at low levels.
As depicted in Figure 1, Paul's disruptive behavior data were variable with a slight upward trend during baseline and during the EFA there was an elevation in the percentage of intervals disruptive behaviors occurred only during the teacher attention condition while maintaining zero to near zero levels during the escape and play conditions, suggesting that attention was the maintaining variable for Paul's problem behaviors. As depicted in Figure 2, the data were elevated and variable for off-task behavior during baseline. Additionally, there was an increasing trend of off-task behavior during the attention condition, while maintaining a low level of off-task behavior during the escape and play conditions, further supporting that attention as the maintaining variable. After the DRO + DRA intervention was implemented, disruptive and off-task behaviors decreased and maintained at zero to near zero levels.
Figure 3 reports the average number of intervals per 5 min session for each condition with teacher attention to the participants' inappropriate and appropriate behaviors. During baseline, higher rates of attention were provided contingent on inappropriate participant behaviors. This contingency was reversed during intervention; higher rates of attention were provided contingent on appropriate participant behaviors. These data provide further support that attention was the maintaining function for the participants' disruptive and off-task behaviors.
[FIGURE 3 OMITTED]
The number of intervals per min with teacher presentation of demands during each condition was variable because the presentation of demands was only programmed during two of the functional analysis conditions. Demand presentation was programmed every 30 s in the escape condition during the functional analysis and only an initial demand was presented during the play condition. Otherwise, demand presentation was allowed to occur naturally during baseline, the attention condition of the functional analysis, and intervention. The highest number of intervals with demand presentation per min occurred during the escape condition. The average number of intervals with demands per min across all participants was as follows: baseline = 0.09, attention = 0.16, escape = 1.84, play = 0.11, and intervention 0.30.
Using a 5-point Likert scale developed by the researchers (1 = strongly agree, 2 = agree, 3= not sure, 4 = disagree, 5 = strongly disagree), data were collected on the acceptability of the current procedures. Lower ratings suggested higher social validity. Four questions were rated by the teacher as strongly agree (rating = 1), three questions were rated as agree (rating = 2), and two questions were rated as not sure (rating = 3). For example, the teacher strongly agreed the procedures for running the functional analysis were easy to learn and perform. He also strongly agreed the procedures for running the intervention were easy to learn and perform. The teacher agreed that the intervention increased on-task behavior, increased the participants' productivity, and decreased inappropriate behaviors. The teacher was undecided if he would recommend using a functional analysis and function-based interventions to other teachers working with students with behavior problems, and was undecided if he would continue using these procedures next year if he had the same students. However, when asked, the teacher reported he responded this way because he had learned the effectiveness of giving attention to appropriate behaviors versus inappropriate behaviors and in the future would employ this specific behavior management strategy before considering the need to run a formal EFA
During the functional analysis, the teacher's overall average percent of 10 s intervals with correct implementation was 97% (range = 80% - 100%). The teacher's overall average percent of 10 s intervals with correct implementation was 99 % (range = 97% - 100%) during the attention condition, 91% (range = 80% - 100%) during the escape condition, and 100% during the play condition. During the function-based intervention, the teacher's overall percent of 10 s intervals with correct implementation across all 3 participants was 99% (range = 97% - 100%). The teacher's overall average across all conditions was 98% with Brandon (range = 90% - 100%), 98% with LaTonya (range = 80% -100%), and 98% with Paul (range = 87% -100%).
Overall, integrity was high across all conditions and observations. The lowest percentage of correct implementation was observed during the EFA escape condition. This was observed to be due to the teacher not providing praise following the participant complying with the academic demand. During the intervention observations, the only observations that were not scores of 100% were due to the teacher providing attention contingent on the participant's disruptive and off-task behaviors. This was recorded a total of three times during all intervention observations.
Results for all 3 participant's EFA's suggested that teacher attention functioned as a maintaining variable for disruptive and off-task behaviors. Based on this hypothesis, a DRO + DRA intervention was implemented providing reinforcement for other behaviors and alternative behaviors (all desirable behaviors), while targeted behaviors were on extinction. This function based intervention was successful in decreasing the rates of disruptive and off-task behaviors for all 3 participants while increasing the rates of appropriate and on-task behaviors.
In the current study, the functional analysis demonstrated that teacher attention maintained disruptive and off-task behaviors for all participants. Based on this hypothesis, a function-based intervention, which included a DRO with extinction and DRA, was implemented. This intervention programmed reinforcement for other behaviors and alternative behaviors, while extinction was implemented for all disruptive and off-task behaviors. This function-based intervention was successful in decreasing the rates of disruptive and off-task behaviors for all participants, in turn resulting in an increase of on-task behavior.
Furthermore, the function-based intervention implemented modified the contingencies that had previously resulted in the participants gaining the teacher's attention. More specifically, the disruptive and off-task behaviors that previously gained attention no longer resulted in gaining the reinforcer, thus, systemically interrupting the re-sponse-reinforcer contingency. Additionally, more socially appropriate behaviors (i.e., on-task behaviors and raising one's hand without engaging in disruptive behaviors) resulted in gaining the preferred reinforcer (i.e., teacher attention).
These results add to the literature by providing further support of the effectiveness of function-based interventions when the function is identified through the use of an EFA. Several researchers have demonstrated larger improvements in behavior when using function-based versus non-function-based interventions (Moore et al., 2001; Wright et al., 2006), and in turn, the use of function-based treatments has decreased the use of punishment-based procedures (Pelios, Morren, Tesch, & Axelrod, 1999). Behavioral interventions that are based on the maintaining function of the problem behavior are more likely to be effective because they modify the maintaining contingency rather than rely on strong reinforcers and/or punishers that override the conditions maintaining the behavior (Mace, 1994).
Moreover, findings from the current study support prior studies using the EFA methodology for students in school settings who have behavioral problems (Ervin et al., 2001), specifically in general education settings (Scott et al., 2004), and with typically developing children (Moore et al., 2001). For example, results from the EFA component matched to other studies reporting that attention is frequently a maintaining function of disruptive classroom behaviors (Ervin, et al., 2001; Lewis & Sugai, 1996; Vollmer & Northup, 1996). Other studies have also found, similar to the current study, that classroom teachers can play a vital role in conducting EFA (Moore et al., 2001; Kamps et al., 2006). In the current study, the EFA provided important information in order to design a function-based intervention using both a DRO with extinction and DRA component.
While the procedures utilized in the current study were effective in the reduction of undesirable behavior for all 3 participants, there is still more that needs to be examined in future studies concerning the teacher's responses on the social validity questionnaire. For example, the teacher reported that the EFA procedures were easy to use and the function-based intervention was effective, but he was undecided if he would recommend or use the procedures in the future. Based on the teacher's verbal report and additional comments made on the survey, the teacher stated that he had learned the effectiveness of giving attention to appropriate behaviors versus inappropriate behaviors and in the future would employ this specific behavior management strategy, before considering the need to run a formal EFA. The teacher anecdotally reported that when he used these strategies with other students, they appeared more focused and on-task during academic times. It is hypothesized that the teacher may have had concerns regarding the efficiency of the EFA procedures similar to conclusions reported by Scott et al. (2004). As elaborated by Scott et al., school personnel may question the efficiency of running an EFA in the regular classroom during normal instruction when there are more students and demands being placed upon the teachers. In the current study, the setting included a small group of students in a reading classroom, in contrast to the majority of general educational settings with larger numbers of children. It is unknown if a teacher with 20 or more students would be able to efficiently and accurately conduct the EFA during normal instruction. Scott et al. extensively discussed the need to make the functional assessment process efficient and effective for teachers, and summarized that functional assessment methods are currently a mix of trial and error in uncontrolled and unmonitored experiments.
Limitations and Future Directions
One possible limitation of the present study was the decision to conduct the EFA conditions during regular reading instruction time. This could have confounded the EFA results because the establishing operation for escape was likely present during all three conditions. The establishing operation was likely present because the participants were instructed to begin reading at the beginning of every class and their reading materials were present during the attention and escape conditions. In addition, during the control/play condition, participants were instructed to read the book of their choice in the reading corner. Although the demand to read was presented at the beginning of the condition, the participants were allowed to pick a high preference book during the control/play condition. During the attention and control/play conditions, scheduled prompts to read were not delivered and reading materials were not removed contingent on disruption. The high level of responding during the attention condition could have occurred because of the presence of the establishing operation for escape, and not because the establishing operation for attention was present. Although this is theoretically possible, results from the EFA suggest that attention was the maintaining variable based for all 3 participants. If the participants were engaging in target behaviors to truly escape the demands and not to gain attention, then the rate of behaviors would have also been high during the escape condition. During the escape condition, attention was provided contingent on on-task behaviors, and was delivered every 30 s. Therefore, it is possible that the establishing operation for attention was also present during this condition, but only for on-task behaviors.
In order to help control for the presence of the establishing operation for escape during EFAs conducted in school settings, future studies could incorporate the escape condition as described in the current study and an escape-to-attention condition as described by Mueller, et al. (2005). This would further differentiate if a participant's escape behavior were maintained by escape from the task versus escape to gain more attention. It may not be possible to completely remove the establishing operation for escape or attention from an EFA conducted in an academic setting, but experimental conditions can be manipulated to enhance our ability to evaluate the variables maintaining a participant's problem behavior within their typical classroom.
Another limitation of the present study is that there were only 3 participants and they were all in the same classroom. To further examine the generality of the procedures used in the present study, replication across multiple teachers, settings, and participants is needed.
An additional limitation of this study is that a DRA in the classroom was used with the intervention, yet the use of the alternative behavior was not measured. One side effect that can arise from using a DRA in a classroom is that students may engage in a high rate of the alternative behavior to gain the reinforcer, causing classroom disruptions (Wright-Gallo et al., 2006). For example, students may have recruited the teacher's attention appropriately by raising their hand on a repeated basis (i.e., unnecessarily), causing a decrease in academic engagement. This did not occur with the current participants, yet future studies should monitor the frequency that the participants engage in the alternative behavior and document if there are any potential negative side effects if the frequency is high.
The current study is only one of a limited number of studies that has evaluated the use of EFA within the school environment and only one of an even smaller number of studies that has evaluated the use of EFA during the student's regular academic instructional activities (Ervin, et al., 2001; Shumate, 2008). Future studies are needed to further evaluate the use of EFA in school settings, more specifically during students' regular academic activities. Additionally, the current study utilized the teacher as the person to implement the experimental conditions and intervention. This is also an area with a dearth of studies and more studies are needed to evaluate the utility of having regular school personnel conducting the experimental conditions and intervention in comparison to a researcher. Furthermore, future research should address whether or not other school personnel trained can successfully conduct the experimental sessions in similar instructional settings (i.e., small group), in instructional settings with larger groups of students (e.g., general classroom instruction, gym class), and across multiple settings. This will allow for an empirical foundation in which to develop a technology of functional analysis for schools to use. Further research is also necessary to determine the efficiency of the teacher conducting all these roles compared to other school personnel.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed. TR). Washington, DC: Author.
Bessette, K. K., & Wills, H. P. (2007). An example of an elementary school paraprofessional-implemented functional analysis and intervention. Behavioral Disorders, 32, 192-210.
Broussard, C., & Northup, J. (1997). The use of functional analysis to develop peer interventions for disruptive classroom behavior. School Psychology Quarterly, 12, 65-76.
Carr, E. G. (1994). Emerging themes in the functional analysis of problem behavior. Journal of Applied Behavior Analysis, 27, 393-399.
Erbas, K., Tekin-Iftar, E., & Yucesoy, S. (2006). Teaching special education teachers how to conduct functional analysis in natural settings. Education and Training in Developmental Disabilities, 41, 28-36.
Ervin, R., Radford, P., Bertsch, K., Piper, A., Ehrhardt, K., & Poling, A. (2001), A descriptive analysis and critique of the empirical literature on school-based functional assessment. School Psychology Review, 30, 193-210.
Good, R.H., & Kaminski, R.A. (Eds.). (2002b). Dynamic Indicators of Basic Early Literacy Skills (6th ed.). Eugene, OR: Institute for the Development of Education Achievement. Available: http://dibcls.uoregon.edu/
Gresham, F. M., Watson, T. S., & Skinner, C. H. (2001). Functional behavioral assessment: Principles, procedures and future directions. School Psychology Review, 30, 156-172.
Kamps, D., Wendland, M., & Culpepper, M. (2006). Active teacher participation in functional behavior assessment for students with emotional and behavioral disorder risks in general education classrooms. Behavioral Disorders, 31, 128-146.
Hanley, G. P., Iwata, B. A., & McCord, B. E. (2003). Functional analysis of problem behavior: A review. Journal of Applied Behavior Analysis, 36, 147-185.
Hoff, K. E., Ervin, R. A., & Friman, P. C. (2005). Refining functional behavioral assessment: Analyzing the separate and combined effects of hypothesized controlling variables during ongoing classroom routines. School Psychology Review, 34, 45-57.
Horner, R. H. (1994). Functional assessment: Contributions and future directions. Journal of Applied Behavior Analysis, 27, 401-404.
Individuals with Disabilities Education Act Amendments of 1997 (PL 105-17). 20 USC Chapter 33, Sections 1400 et seq.
Individuals with Disabilities Education Improvement Act 2004, 20 USC [section] 1400 et seq. (reauthorization of the Individuals with Disabilities Education Act of 1990).
Iwata, B. A, Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1994). Toward a functional analysis of self-injury. Journal of Applied Behavior Analysis, 27, 197-209. (Reprinted from Analysis and Intervention in Developmental Disabilities, 2, 3-20, 1982).
Iwata, B.A., Wallace, M. D., Kahng, S., Lindberg, J. S., Roscoe, E. M., Conners, J., Hanley, G. P., Thompson, R. H., & Worsdell, A. S. (2000). Skill acquisition in the implementation of functional analysis methodology. Journal of Applied Behavior Analysis, 33, 181-194.
Lewis, T. J., & Sugai, G. (1996). Descriptive and experimental analysis of teacher and peer attention and the use of assessment-based intervention to improve pro-social behavior. Journal of Behavioral Education, 6, 7-24.
Mace, F. C. (1994). The significance and future of functional analysis methodologies. Journal of Applied Behavior Analysis, 27, 385-392.
Mace, A. B., & West, B. J. (1986). Analysis of demand conditions associated with reluctant speech. Journal of Behavior Therapy & Experimental Psychiatry, 17, 285-294.
Moore, J. W., & Edwards, R. P. (2003). An analysis of aversive stimuli in classroom demand contexts. Journal of Applied Behavior Analysis, 36, 339-348.
Moore, J. W., Edwards, R. P., Wilczynski, S. M., & Olmi, D. J. (2001). Using antecedent manipulation to distinguish between task and social variables associated with problem behaviors exhibited by children of typical development. Behavior Modification, 25, 287-304.
Mueller, M. M., Sterling-Turner, H. E., & Moore, J. W. (2005). Towards developing a classroom-based functional analysis condition to assess escape-to-attention as a variable maintaining problem behavior. School Psychology Review, 34, 425-431.
No Child Left Behind Act of 2001, Pub. L. No. 107-110, 115 Stat. 1425. (2002).http://www.ed.gov/legislation/ESEA02/
Pelios, L., Morren, J., Tesch, D., & Axelrod, S. (1999). The impact of functional analysis methodology on treatment choice for self-injurious and aggressive behavior. Journal of Applied Behavior Analysis, 32, 185-195.
Sasso, G. M., Reimer, T. M., Cooper, L. J., Wacker, D., Berg, W., Steege, M., Kelly, L., & Allaire, A. (1992). Use of descriptive and experimental analyses to identify the functional properties of aberrant behavior in school settings. Journal of Applied Behavior Analysis, 25, 809-821.
Scott, T., Bucalos, A., Liaupsin, C, Nelson, C. M., Jolivette, K., & DeShea, L. (2004). Using functional behavior assessment in general education settings: Making a case for effectiveness and efficiency. Behavioral Disorders, 29, 189-201.
Scott, T. M., & Kamps, D. M. (2007). The future of Functional Behavioral Assessment in School Settings. Behavioral Disorders, 32, 146-157.
Shumate, E. D. (2008). School-based Experimental Functional Analysis of Problem Behavior: A Review. Unpublished Manuscript, University of Kansas, Lawrence, KS.
Sterling-Turner, H. E., Robinson, S. L., & Wilczynski, S. M. (2001). Functional assessment of distracting and disruptive behaviors in the school setting. School Psychology Review, 30, 211-226.
Sullivan & Buchanan (1988). Sullivan's Programmed Reading. Phoenix Learning Resources, Inc. New York.
Vollmer, T. R., & Northup, J. (1996). Some implications of functional analysis for school psychology. School Psychology Quarterly, 11,76-92.
Wright-Gallo, G. L., Higbee, T. S., Reagon, K. A. & Davey, B. J. (2006). Classroom-based functional analysis and intervention for students with emotional/behavioral disorders. Education and Treatment of Children, 29, 421-436.
This research was supported in part by a Grant (H324X010011) from the Office of Special Education Programs (OSEP). We thank Debra Kamps for her comments and suggestions on earlier drafts of this manuscript. Finally, we would like to thank the teacher and students who participated in the study.
Correspondence to Howard P. Wills, Juniper Gardens Children's Project, 650 Minnesota Avenue, 2nd Floor, Kansas City, Kansas 66101-2800; email: firstname.lastname@example.org.
Emily D. Shumate
Howard P. Wills
Juniper Gardens Children's Project
University of Kansas