Students who engage in challenging behaviour are often restricted from accessing social and educational opportunities Bradley, Henderson, Monfore, 2004; Wehby Lane, 2009. Up to 20% of students who engage in challenging behaviour do not respond to universal behaviour management strategies (Kamps et al., 2011; Sugai & Horner, 2009; Warren et al., 2003). At least 5% of students will require individualised behaviour support strategies based on an understanding of the behavioural function (Sugai & Horner, 2009; Walker, Cheney, Stage, Blum & Horner, 2005; Warren, et al., 2003). Students that repeatedly use challenging behaviour, do so to have a need met, these are often triggered by antecedents and maintained by reinforcing consequences (Carr et al., 2002; Cooper, Heron & Heward, 2007; Iwata & Worsdell, 2005; Sugai & Horner, 2009). However, in the complex, fast paced and unpredictable environment of a classroom, teachers rarely have the opportunity to consider what need is not being met (Emmer & Stough, 2001), often applying reactive punitive approaches (Scott, McIntyre, Liaupsin, Nelson, Conroy & Payne, 2005; Van Acker, Boreson, Gable, & Potterton, 2005; Wubbels, 2011), or making vague and fruitless attempts at implementing strategies unrelated to factors maintaining the behaviour (Blood & Neel, 2007; McIntosh, Brown, & Borgmeier, 2008; Scott et al., 2005).
Positive behaviour support (PBS) is an approach used by many researchers and practitioners with aims to enhance participant quality of life and in doing so reduce the participant's need to use challenging behaviour (Carr et al., 2002). Since its inception in the 1980's, research in PBS has illustrated the effects of increasing positive behaviour using non-aversive interventions and as a result reducing challenging behaviour (Carr et al., 2002). One of the core principles of PBS is the use of a functional behaviour assessment (FBA) when supporting students that require individualised interventions (Hieneman, Dunlap, & Kincaid, 2005; Sugai & Horner, 2009; Walker et al, 2005; Warren et al., 2003; Tarbox et al., 2009). The purpose of the FBA is to understand the function of the behaviour, beyond what the topography can provide (Iwata & Worsdell, 2005; Kates-McElrath et al., 2007; McIntosh et al., 2008; Scott, Anderson, & Spalding, 2008; Tarbox et al., 2009). In PBS, the FBA allows for an individualised intervention or strategy to be developed which works to modify the individual's environment to encourage more pro-social behaviour (Carr et al., 2002; Iwata & Worsdell, 2005; Koegel et al., 2012; Sugai & Horner, 2009; Umbreit, Ferro, Liaupsin & Lane, 2007). The pro-social behaviour is identified to meet the functional equivalence of the challenging behaviour, rendering the challenging behaviour irrelevant and unnecessary for the student to use (Carr et al., 2002). Therefore a FBA can equip a teacher with a fundamental tool that can be applied to many students on an as need basis. This is particularly important as there is insufficient availability of specialists that are skilled in conducting FBAs (Grey, Honan, McClean, & Daly, 2005) even more so in at least some parts of Australia than in the United States (Mooney, Dobia, Barker, Power, Watson, & Yeung, 2008).
There are three forms of FBA commonly reported in the literature. Indirect methods (e.g. interviews, rating scales, file reviews) (Alter, Conroy, Mancil, & Haydon, 2008; O'Neill et al., 1997; Tarbox et al., 2009) direct or observational methods (e.g. ABC Analysis) (Alter et al., 2008; Bijou, Peterson, & Ault, 1968; McIntosh et al., 2008), and experimental functional analysis (FA) (McDonald, Moore, & Anderson, 2012; Ellis & Magee, 2004; Iwata, Dorsey, Slifer, Bauman, & Richman, 1982 / 1994; Iwata & Worsdell, 2005; Sigafoos & Sagger, 1995).
Validity and feasibility concerns exist regarding all three methods. Indirect methods are fallible as they rely on memory recall and the ability of the participant to observe antecedents and consequences (Alter et al., 2008; Tarbox et al., 2009; McDonald et al., 2012). Direct observation requires an observer who is available and attending to the target student at all times in order to keep records of the student's behaviour, antecedents and consequences; however in the busy classroom it is difficult to observe antecedents and consequences. To overcome these difficulties, studies have demonstrated the applicability of video technology (Behavior Imagingâ„¢) in conducting direct FBA (McDonald et al., 2012; Reischl & Oberleitner, 2009). Behavior Imagingâ„¢ is a camera and computer software program that is accessed via a remote control to video record behavioural incidences, antecedents and consequences as they occur in the natural environment (Reischl & Oberleitner, 2009). Although the experimental FA provides the most tangible results due to its experimental cause and effect nature (Iwata & Worsdell, 2005), it is traditionally a lengthy process and not suitable for the everyday classroom (Iwata, Dorsey, Slifer, Bauman, & Richman, 1982 / 1994;). However, some recent studies have illustrated the feasibility of an adapted brief version (McDonald et al., 2012; Sigafoos & Sagger, 1995).
Research indicates that many use indirect approaches in the field (Blood & Neel, 2007; O'Neill & Stephenson, 2010; Machalicek, O'Reilly, Beretvas, Sigafoos, & Lancioni, 2007), research suggests that these methods are not as valid as direct assessment methods (Alter et al., 2008; Tarbox et al., 2009), and some have found that only the experimental FA provides the most relevant and valid results (McDonald et al., 2012).
Although a lot of research exists to support the efficacy of using PBS practices in the classroom (e.g., Grey et al., 2005; McIntosh, et al., 2008; Stage et al., 2008; McDonald et al., 2012), some research has shown that teachers who have been trained, do not always apply their training in the field (Blood & Neel, 2007; Boardman, Argüelles, Vaughn, Hughes & Klingner, 2005; Scott et al., 2005; Smith, Richards-Tutor & Cook, 2010; Van Acker et al., 2005). Van Acker et al. (2005) found that after they provided training in positive behavioural support practices, teachers continued to use inappropriate methods in conducting an FBA and were unable to select appropriate replacement behaviours. Blood and Neel (2007), in their study identified that none of the six participating teachers were able to identify the function-based intervention that had been written in students' intervention plans. In another study, Blood and Neel (2007) reviewed 43 student files that had included an FBA, however the intervention selected most often did not seem to match the assessed behavioural function. These studies highlight the research to practice gap that have been noted by many (Boardman et al., 2005; Burns & Ysseldyke, 2009; Cook & Cook, 2011; Couvillon, Bullock & Gable, 2009). These studies however do not identify why the gap exists.
Social validity measures were developed in the late 1970's to gain insight into the participant's perspectives and feasibility of the practices and are considered a crucial part of single subject design (Carr et al., 2002; Horner, Carr, Halle, McGee, Odom, & Wolery, 2005; Machalicek et al., 2007; Schwartz & Baer, 1991), however the high number of articles published with positive feedback and high social validity does not shed much light on what may be contributing to the research to practice gap (Machalicek et al., 2007). As one author notes, social validity should be identifying what may be preventing a process from working rather than re-illustrating how that process does work (Schwartz & Baer, 1991), then trainers and practitioners may be better able to predict when teachers may fail to implement PBS practices and how best to support them to overcome the potential barriers of adopting PBS practices. While single subject research allows for adaptability and flexibility, most studies have been primarily guided by researchers who have specific training in the area, and who do not always consider the perspective of the teacher (Kates-McElrath, Agnew, Axelrod, & Bloh, 2007; Machalicek et al., 2007). This may be crucial in identifying approaches and strategies that will be implemented rather than ones that will be avoided by the teacher. This may also shed light on some of the research to practice gap that exists in PBS practices.
Some studies that have explored teachers' perspectives, although informative are limited as they have not been conducted alongside the experimental and hands on components of PBS practices. Findings from teacher perspectives research has found that when implementing behaviour management strategies teachers report: ease of implementation, the personal feelings they hold, the beliefs about the appropriateness for the student and classroom as well as the availability of support staff and materials all play an important role (Boardman et al., 2005). In addition teachers have a preference for on the job collegial training and support (O'Neill & Stephenson, 2010).
The purpose of this study was to conduct a mixed method design case study in a classroom with a student engaging in hair pulling behaviour. Experimental analysis of FBA and PBS processes was conducted and paralleled a qualitative analysis of a teacher's experience of these processes. The focus was to understand the barriers and facilitators towards a teacher using these practices in a classroom, in hopes to further understand the research to practice gap. Three methods of FBA were used to identify the function of behaviour. A single subject AB design was used to investigate the functional relation between teacher selected strategies prior to and post identifying the behavioural function. Qualitative case study methods were used to gain the participating teacher's perspective of (1) understanding student problem behaviour, (2) the process of conducting a FBA in the classroom, (3) strategy planning and selecting an appropriate intervention, and (4) what additional needs are required to support a student with problem behaviour in the classroom.
In accordance with the research literature it was hypothesised that the direction observation and experimental FA would yield the most accurate results of behavioural function; and a function-based intervention would result in the reduction of hair pulling behaviour.
Prior to commencement, this study was approved by Monash University's Standing Committee on Ethics in Research Involving Humans and the Victorian Government Department of Education and Early Childhood Development Ethics Committee. The participating school and teacher provided informed consent to be involved in the study. Informed consent was obtained from the participating student's parents on the student's behalf.
Participants and Setting
The participating teacher, pseudonym Melissa, was employed at a special developmental school and was a teacher to a grade 1 equivalency class. Melissa sought assistance to address a specific student's behaviour of pulling other students hair in the classroom. This student also served as a participant.
Melissa was a teacher with over 25 years of experience. She had primarily conducted her training in the late 1970's and was certified as a kindergarten teacher. She had no formal training in special education; however she had been working in special schools for most of her career. She was employed at secondary special education school for 11 years and moved to her current early years special developmental school 10 years ago. Melissa reported that she had no experience with Functional Behaviour Assessments (FBA).
The participating student, pseudonym Joe, was a 7-year-old boy who was diagnosed with autism spectrum disorder. Joe had attended the school for almost two years. However hair pulling behaviour had only started four weeks prior to the commencement of this study. Joe was vocal and could verbalize simple sentences, in a non-conversational manner. Joe could follow complex two-step instructions.
The study was conducted in the special developmental school during regular scheduled classroom activities. The classroom was staffed by one teacher and two teacher assistants. The class was made up of seven students, all of whom had been diagnosed with a developmental delay and / or intellectual disability.
The study used both quantitative and qualitative procedures to gather data.
Quantitative Measures and Design
Functional behaviour assessment (FBA) materials.
This study utilised three types of measures in conducting a functional behaviour assessment: The Functional Assessment Interview Form (FAI) (O'Neill et al. 1997); an Antecedent Behaviour Consequence (ABC) direct observation worksheet (Bijou et al., 1968); and a condition specific experimental functional analysis interval recording sheet developed for this study.
The FAI (O'Neill et al. 1997) is an indirect assessment of behavioural function. It is a structured interview form used to guide a respondent through describing behaviours, antecedents, consequences and setting events that may be involved in predicting or maintaining the problem behaviour. A modified version of the FAI (See Appendix x) was used to develop hypotheses regarding possible functions for the problem behaviour. The study also used a modified version of the ABC observation worksheet (See Appendix x) as a direct form of FBA. Three columns are readily used in an ABC observation worksheet, antecedent, behaviour, and consequence are listed at the top (Bijou et al. 1968), and this study included a fourth column, precursor behaviour to identify the participating student's behaviour just prior to engaging in hair pulling behaviour. From the observations made using the ABC worksheet, patterns of antecedents, behaviours and consequences emerge to provide a hypothetical function for the problem behaviour. All incidences of hair pulling behaviour were recorded by means of Behavior Imagingâ„¢ technology, which captured video footage of 2-minutes prior to and 2-minutes post the hair pulling incident. These recorded digital files were viewed and ABC data worksheets were used to record events.
The experimental functional analysis (FA) required the direct manipulation of materials naturally found in the classroom and the cooperation of a peer. An event recording sheet was developed to record the presence or absence of problem behaviour for each trial (Adult attention, access to directive, and peer demand) or control condition (See Appendix x).
Intervention selection materials.
The function based intervention decision model (Umbreit et al., 2007, see Appendix x) was used to guide the teacher's selection of an appropriate functionally equivalent replacement behaviour that the student could use to replace the hair pulling behaviour. The model represents a direct link between the results of a FBA and a successful intervention that supports a student's use of socially appropriate replacement behaviour. The three main methods of intervention used in this model include: 1) teaching the student replacement skills and building general competence, 2) making environmental adjustments, and 3) adjusting contingencies.
Data collection procedures.
Frequency data was collected on hair pulling behaviour during the regular classroom routine. Data was only captured when the student was present in the classroom and on days in which the student was present for an entire day. Data was not captured on days in which the student spent drastically less than the usual amount of time (4.5 hours) in the classroom i.e. for days that included lots of outside activities. Data were collected 3 days per week and was video recorded by either the classroom teacher or the researcher by means of Behavior Imagingâ„¢ technology (Reischl, & Oberleitner, 2009) and the number of incidences per day was tallied, based on the incidences recorded. The Behavior Imagingâ„¢ system consisted of a laptop with Behavior Imagingâ„¢ Capture software, a webcam and remote control button. The Behavior Imagingâ„¢ system was set up in the classroom and the teacher was taught how to use it. The teacher was instructed to activate the system every time the target behaviour occurred in the classroom, by pressing the remote button.
Due to the concerning nature of hair pulling behaviour on other students' wellbeing, the teacher was advised to prevent the behaviour where possible and record the incidence. Hair pulling behaviour was prevented 22 out of the 34 times it was recorded as an incident.
Dependent variable and Research design.
The primary dependent variable for this study was pulling other student's hair. The experimental procedure was a quasi-experimental single-subject AB design. The design had two phases: (a) baseline, in which the teacher conducted her class in a routine manner and implemented consequence based time out behaviour management strategies for the problem behaviour; and (b) intervention, in which the teacher collaborated in and selected appropriate strategies to be implemented. The FBA was conducted during baseline.
Interobserver agreement (IOA).
Interobserver agreement data was collected for 33% of videos used in the ABC analysis and 33% of videos documenting the experimental Functional Analysis (FA) trials. Interobserver agreement for the FBA data was calculated using an interval-by-interval procedure with a second, trained, observer independently viewing the recorded sessions and comparing these results with those of the primary observer. Total interobserver agreement was calculated by dividing the number of intervals of agreement within each interval by the number of intervals of agreement plus disagreement, multiplied by 100 (Cooper et al., 2007). IOA measures of the ABC analysis videos provided three results, an IOA mean of 66.63% for antecedent events; an IOA of 100% for the problem behaviour; and an IOA mean of 94% for the consequential events. IOA measures of the FA trials indicated 100% agreement on the occurrence or non-occurrence of problem behaviour on all trials.
Qualitative Design and Measures
Research methods and data collection.
Qualitative case study methods were used in which the phenomena of interest were Melissa's perspective of: (1) understanding Joe and his hair pulling behaviour, (2) the process of conducting a FBA in the classroom, (3) strategy planning and selecting an appropriate intervention, and (4) additional needs required to support a student with problem behaviour in the classroom.
The case study may be characterised as an intrinsic case study (Stake, 1995) in that the teacher, Melissa's perspectives were used to explore the particular case of using FBA and intervention selection practices in this classroom for this student. Qualitative data were gathered within a five week period, which coincided with the phases of the single-subject research design. The primary methods used were semi-structured interviews with Melissa as the informant and a written logbook was kept by the researcher to keep an audit trail of the process. The first interview was conducted during the baseline phase and consisted of questions relating to teacher's beliefs of Joe and his behaviour and strategies used by the teacher. The second interview occurred shortly after the baseline phase, once all FBA procedures were complete and focused on Melissa's perspectives on conducting the FBA and how the FBA suited her, the student and the classroom environment. In addition the second interview focused on the process of selecting an intervention. The third interview occurred at the end of the researcher's involvement, with a focus on the teacher's perspectives and reflection of the practices of FBA and PBS and the supports needed. Each interview took between 30 and 60 min and was conducted in the teacher's classroom. The interviews were audio recorded. Prior to data analysis, interview audio files were transcribed. Questions were identified by the research team to guide discussion (see Appendix x).
Qualitative data analysis.
At the completion of the study, the researcher coded the transcribed interviews. Initially, an open coding approach (Liamputtong & Ezzy, 2007) was used to compare and analyse the qualitative data. Preliminary categories that were found to be repeated in the data were defined and used to guide the initial coding of the data. In total, seven categories emerged from this process. These categories were defined in terms of their properties and dimensions and the data were sorted according to these categories. Next, axial coding was used to re-examine the categories identified to determine how they may be related or linked to one another as well as the core issues that the study aimed to identify. Through an inductive and deductive process, a final set of categories were proposed that were observed to exist in the data collected. Four main categories or themes emerged upon final analysis.
Member checks. To enhance the internal validity of the findings, Melissa was given the transcripts of each interview and was provided with the opportunity to edit any information within them. Melissa chose not to make changes to the data.
Functional behaviour assessment (FBA).
The FBA consisted of three phases. Phase 1 included a teacher interview using a modified FAI to obtain as much information as possible to inform and develop a hypothesis of function. Due to the availability of the teacher and unforeseen circumstances, this interview was conducted seven days after the commencement of baseline data collection. Phase 2 included the direct FBA which used the ABC direct observation recording worksheets. Video footage was viewed by the researcher after the FAI had been conducted to collect antecedent, behaviour and consequent data for each occurrence of the behaviour. On one occasion the teacher was involved in viewing four of the videos and contributed to the process of identifying the antecedents and consequences to the hair pulling behaviour. In the third phase of FBA an experimental functional analysis was conducted to identify the variables that maintained Joe's hair pulling behaviour. The discrete trial functional analysis methodology was as described by Sigafoos and Saggers (1995). Over two days Joe was exposed to three assessment conditions: (a) access to adult attention, (b) access to instruction, directive or activity, and (c) escape from peer demand, with up to five trials under each condition. Each trial took 2-minutes (60 s each for the trial and control conditions) and trials were distributed randomly throughout the day. A couple of trials of task demand were also trialled by the teacher. Following these three phases a collaborative meeting was held with the teacher and researcher. Strategies and function-based interventions were discussed for implementation.
Intervention selection procedures.
The results of the FBA indicated the function of Joe's hair pulling behaviour to be maintained by a peer. This knowledge resulted in environmental changes implemented by the teacher and teacher assistants. The teacher intervened with the peer, moved Joe's seat, provided emotional support and regularly checked in with Joe, thus resulting in a reduction in Joe's hair pulling behaviour.
Due to the concerning nature of hair pulling behaviour on other students, and the immediate reduction in hair pulling behaviour without experimental control, this study did not follow a strictly experimental design to test intervention. The teacher was guided through Umbreit et al.'s (2007) model; and collaboration between the researcher and teacher identified replacement behaviour for implementation. However due to the nature of the behaviour and the case it was determined that a primarily environmental approach to PBS would be utilised and Joe would be better able to access learning if he was moved to another, age appropriate classroom. Although no intervention for replacement skill development was implemented, the teacher discussed it as a possibility in the future.
Functional Behaviour Assessment Results
Results of three phases of the FBA are provided: The Functional assessment interview, Antecedent, behaviour, consequence (ABC) analysis, and experimental functional analysis (FA)
In the FAI the teacher reported that Joe engaged in hair pulling behaviours daily and the teacher believed this to be problematic and disruptive throughout the school day. Joe's teacher, Melissa, reported that no identified cause was known. Melissa hypothesised that the behaviour may be related to environmental noise levels being too loud. Melissa identified that the consequence that was in place, time out, was not appropriate as it did not reduce Joe's hair pulling behaviour. However she could not identify how time out might be providing Joe with an outcome. Melissa reported that Joe might engage in the behaviour if he was left alone for 15 minutes, however was not sure. Joe's favourite items were reported to be cars, trains, i-pad, tracing activities, picture books and DVD covers.
During the FAI, Melissa did not recognise the possibility of other students affecting Joe and his hair pulling behaviour. However, after some classroom observation and more detailed history, it began to emerge that his hair pulling behaviour could be related to a specific student in the classroom. This particular student had a history with Joe, in which she had made high demands of him, and although these demands had stopped, Joe appeared apprehensive of this peer on many occasions. This history helped in the final two phases of the FBA.
A total of 34 incidences of hair pulling or attempts to hair pull were recorded by the Behavior Imagingâ„¢ capture system, 33 of these videos were used in an ABC analysis. Figure 1 shows the results of the ABC analysis. In the 33 analysed videos, a variety of factors appear to be contributing to the problem behaviour. Low to no attention from the teacher preceded hair pulling behaviour on 14 occasions, the specific peer was present prior to the hair pulling behaviour on 23 occasions, Joe was not engaged in any activity prior to the hair pulling behaviour on 10 occasions. Regarding the consequences to the hair pulling behaviour, these data indicate that on 23 occasions the behaviour was followed by time out, which includes some form of teacher attention and removal from peer, while on eight occasions the consequence was provided in the form of teacher attention and a direction to an activity (no time out). Joe was seen twice to relocate and remove himself from his peer after pulling another student's hair.
Figure 1. Results of ABC analysis using data generated by the Behavior Imagingâ„¢ capture software.
*refers to a identified student being present, and spontaneously appears, yells out, talks to or touches Joe.
These results indicate that Joe's hair pulling behaviour was maintained by negative reinforcement in the form of escape from a peer and / or positive reinforcement in the form of access to adult attention.
The result of the FA, which was conducted over two days, is shown in Figure 2. Possible attempts to hair pull behaviour occurred two out of five times in the access to directive condition and no times under either access to attention or peer demand. Access to attention was trialled five times. Peer demand was only trialled three times due to the nature of the trial, in which the peer was requested to make a demand of the student. However, during all three occasions the peer sat next to Joe, followed the request of the researcher and both students were provided with full adult attention. No hair pulling or attempts occurred in any of the control conditions.
Figure 2. Number of hair pulling incidences during the FA trial or control condition across each of the three conditions.
These results indicate that Joe's low rate of hair pulling behaviour was maintained by positive reinforcement in the form of access to a directive or activity.
Results of hair pulling frequency before and after FBA
Analysis of the frequency of hair pulls engaged by Joe decreased as a result of the teacher's knowledge of the function of behaviour after conducting the FBA. Function based environmental changes were implemented that are likely to have contributed to the reduction in hair pulling behaviour. Hair pulling behaviour was observed at least twice a day and up to eight times a day before the function of the behaviour was known. This reduced to zero times for most days and only once was the behaviour observed after this time. At follow up two weeks later, the teachers reported that the student had not engaged in any hair pulling behaviour in the classroom since the research project had ceased.
Figure 3. Frequency of classroom hair pulling behaviour in incidences per day
Semi-structured interviews with Melissa, Joe's teacher, provided qualitatively rich data parallel to the phases of the FBA and intervention selection procedures. The data collected provided insight into four core issues on four core issues: (1) understanding students' behaviour, (2) the process of conducting a FBA in the classroom, (3) strategy planning and selecting an appropriate intervention, and (4) additional needs to support a student with problem behaviour in the classroom. The information was collected during three phases of the project prior to, during and post the identification of the function for Joe's hair pulling behaviour.
Understanding students' behaviour.
Before Melissa commenced the FBA process, she discussed several reasons for why students behaved as they did, seemingly without understanding the core function. When attempting to explain Joe's behaviour, Melissa reports:
Initially I thought it was anxiety when the noise level increased in the classroomâ€¦ But then I thought well, as well as that he's an opportunist,... So I don't knowâ€¦ initially I thought that, so I don't know whether when he was using that opportunity whether there was noise volume as well. But now it could be an attention thing, I'm changing my mind.
At the second interview, Melissa's understanding of Joe's hair pulling behaviour had changed and had become quite concrete:
I do think it's because of [the identified student] manipulating him and [being] in his personal space, ... he just feels extremely anxious and it is obvious when she makes a move towards him, his behaviour changes. ...Just releasing his tensionâ€¦, it's cause effect type behaviour.
Melissa began to indicate that all the students in her classroom were misbehaving:
... they're all very knowing little people and they pick up very quickly ... they see the limits not being set, they take advantage...
I'm almost thinking that [the identified student's] behaviour is just because she's stubborn and she just wants her own way... just, that's her
In the final interview, Melissa discussed the function of Joe's hair pulling behaviour:
I think we really did get to the core problem... [the identified student] sliding things away from Joe and getting in his space.... I think the hair pulling focused all our attention on him so he felt safer. ... [Joe needed] to seek an adult in [his] environment for help. ... But until he can actually do that, he's not independent to cope with these issues.
Melissa reported that Joe's behaviour had changed since he was moved to another age appropriate classroom:
He's so excited to be going into the [other classroom]. And yesterday, he actually said hello to me. So he's slowly starting to work out that, okay I'm here but I can still talk to Melissa, and .... the other kids ... I can still be their friend.
The process of conducting a FBA.
During the initial interview, no form of functional behaviour assessment (FBA) had yet been conducted; however the Behavior Imagingâ„¢ system had been set up and was recording the incidences of hair pulling behaviour, to be later used for ABC analysis. Melissa discussed her views towards these videos:
There's so much going on in a classroom..., he starts doing his own thing, it's really hard to observe everything. So the video will be good for that.
By the second interview all three forms of FBA (indirect, direct and FA) had been conducted. And Melissa was at least in part involved in each method. Melissa discussed her views of observing the behaviour and antecedents:
our day is to teach and guide and instruct.... Within the hustle and bustle of the day, because it's always a very busy day, just to take that step back and do a bit more observation. I do observe but perhaps not in as much depth as this.... So [the video] it's probably a good way to do it.
Melissa advised she found analysing the videos alongside the researcher to have been the most effective and insightful in identifying the function of Joe's behaviour.
I would be more inclined to think the videos and discussions with the videos. ... the videos they've been really, really effective in finding out the reasons why Joe has behaved the way he has been.
Melissa found the experimental FA trials the most difficult to implement.
it was a little bit difficult, it's easier if you're in a room doing the trials, because we're busy with the other children as well.... it is difficult for us to do trials.... whereas under normal test - a test situation he wouldn't be in a classroom environment. So I think that is quite a difficult thing to expect to happen in a classroom.
Melissa was fairly confident that there was only one function for the behaviour and justified why the behaviour might look like other forms of behaviour:
I do think it's because of [the identified student] manipulating him and in his personal space. ... I think that's the underlying reason, now from observing. I think when he's not engaged in an activity he's probably more aware of what [the identified student] is doing, because that's when he starts scanning the room.
However, she was open to conducting a couple of trials using the experimental FA approach to rule out lack of engagement.
I'll have a look. ... But yes I would be very interested to see what the outcome of not having an activity and not being engaged, what impact it did make on Joe.
When asked if Melissa would conduct an FBA in the future she stated:
Yes I would, yeah, only I wouldn't have access to a video, but ... yes it would be a lot more analysing.
When asked if she would use it with the other identified student, she replied:
Well possibly, possibly I could, but I almost feel that [the identified student's] behaviour has a lot to do with just, that's her.
At the final interview, Melissa weighed up the three FBA methods and advised that the study had been valuable and worthwhile:
Yeah, very worthwhile ..., being able to ... watch that video told us a lot that we hadn't actually picked up in the chaos of the day. ... So, yeah I found that fantastic. ... you do reflect and you analyse ... but perhaps not quite the same as the video shows us... And I think, too, sometimes when children are passive and they sit and they're doing what they've been asked to do, you do tend to help the ones that are more unsettled ... [Video was easier] because that happened in the normal running of the classroom.... From the videos ...Well then I focused more ... Just highlighted different things that we needed to look at.
Although Melissa had agreed to revisit and conduct a couple of FA trials after the previous meeting, she never did. She stated:
Well I think we really did get to the core problem. ... Yeah, but no, I never revisited that. ... it was really difficult. The way the room was going, to work one-on-one or ... to set up something and ... to be constantly watching his behaviour, it's difficult in the mix of a classroom running.... But, I don't know, for some reason it just seemed very difficult... it was just another task that just got too hard.
Strategy planning and selecting an appropriate intervention.
At the initial interview it was clear that Melissa relied most on removal and reactive strategies:
Something has to be done; he just can't carry on his day without a consequence of some sort. ... He was being time outed ... it removes him from the group ...[so] all the children feel safe in the room. ..... But it's not effective-
In addition to time out, Melissa occasionally used pictured sentences to remind Joe after he pulled hair 'I keep my hands to myself', 'I do not pull hair'.
We've got the sentences ... So it's giving him something to do with his hands, rather than hair pulling. ...We made a social story about [another behaviour exhibited on the bus], and the parents would read it, and we would read it and by the end of it, he was just beside himself ... it didn't stop the behaviour. ... So I don't know why I did those sentences now I'm re-thinking it, because that didn't work.... I guess it was just something to focus him on, it was something to trial.
Melissa discussed a proactive and positively focussed strategy:
I just want to try the, Let's Make A Deal Strategy which focuses on positive behaviour and positive outcomes. ... So that we're moving away from the negative behaviour and we're trying to encourage good behaviour. .... I would give him a star for ... really good behaviour. ... And I would keep reinforcing, so that he's continually focussed on ... tasks and then once he'd reached 3 stars maybe to begin with, then he would get his reward [bike ride, trampoline, something really special]... It's a very intense program.
It was evident that Melissa occasionally relied on hope, particularly in the early interviews.
I was hoping that it would fade out over the holidays. I was hoping he'd come back without that behaviour... I'm hoping that [to] lead him in a more positive way... which hopefully would stop him thinking, oh I'm going to get up and pull [another student's] hair. ... I would still expect the hair pulling to happen, at the beginning of the program, but I would hope it would ease off.
At the second interview, Melissa discussed her focus and observations for behaviour management and strategy development:
We decided that we would have absolute zero tolerance for any inappropriate behaviour. ... Consequently [one student] was taken out of the room 3 times. ... we had the behaviour again. So she was removed again. And I decided that it wouldn't be her choice to come back in. It would be on my terms not hers. And I think that really had an impact on her because the third time she was removed she stopped and thought about it ... She responded really well. And then we noticed that she was holding herself back as the day moved on.
In regards to Joe's behaviour management, Melissa discussed her approach and observations:
in that interim the behaviour had increased, because it wasn't stepped on straight away... So setting those boundaries for [the specific peer] showed Joe that his environment probably was becoming a lot safer ... yesterday, he was feeling much more secure, because we were following through consequences for [the other peer], and he was observing but not reacting. .... it's the most effective [strategy]
he has become quite attached to that cars book. ... So I think that ... helps him cope with the situation that's happening in the room if he's got something to physically hold. ... I don't know ... I think it helps him feel more secure, more safe to have something tangible to actually physically hold, ... while he's watching us deal with [the identified student's] behaviour.
Melissa discussed the outcome of the 'let's make a deal strategy':
[this strategy] is not really appropriate for him ... With other children who behave inappropriately it's to try and get the child behaving and working and cooperating while working for that favourite reward ..., for Joe that's not quite the objective. ... to make him work for something and keep him focused on something and really stress that he's got to get that star and do different tasks while [the specific peer] is behaving inappropriately it's not, that's not fair on him.
After talking through Umbreit et al.'s decision making model, Melissa started to discuss alternative strategies for the future:
Maybe he could ask for time just to remove himself from the situation. So learning the skills of requesting a break from the room. .... But initially it would be us teaching him each one. ... so that he can actually request to step outside and calm, away from the stress that he's feeling
By the final interview an environmental modification strategy was implemented:
[Joe's hairpulling] did sort of ease back a lot. And that's probably - well I felt that I was trying to keep [the specific peer] at a particular distance so that she couldn't interfere with him. And also spend time with him, but then we worked out ... So that book helped and the DVD cover. ... and then I moved him to the other side of the table....we've got to be very aware of [the identified peer] and Joe at the same time
We needed to find some way of helping him calm down ... So we decided that Joe should be removed from the class to help him have happier days and not be so stressed. ... We could concentrate a little bit more on the [other students] and we've calmed down because we - I felt like we were all on high alert.
Melissa advised that the current strategy of moving Joe to another room was a 'quick fix' and that more a more detailed intervention should be planned for the future:
he needed to learn strategies when feeling scared and unsafe. That he needs to find a staff member before he tries to deal with it himself in inappropriate ways... I'm just thinking Joe's ... a very passive little boy - so he could be a victim down the track. ... But it's always a slow process to teach a child something like that and it needs a lot of staff input. ... So it's definitely a skill and I actually did have that as a goal for him for this year. But obviously not enough work was done ... as formally as maybe we would do it now because it's quite a serious thing.... so it can be a future recommendation.
Melissa discussed the results of the time out strategy she had implemented with Joe:
I did the time out because he had to see that there was a consequence for what he did. He had to know that that was wrong. And as well as that, the other children had to see that Joe was given a consequence. Otherwise, that's not fair in their eyes ... But now that we know the reason for it, I think to hopefully intervene or, and redirect would be the way to go rather than sitting him out ... I knew that that wasn't effective, but it was just something that had to be done at that time until something was worked out. .... Well [other students] could start being ridiculous and doing the same sort of thing, thinking there are no rules, no consequences in this room, I'll just run amok ... So I'm hoping that because they saw something was done about that behaviour, that they didn't think okay they can just willy-nilly pull hair too.
When asked if she would use time out again with another student like Joe, Melissa responded:
Probably, yeah to begin with, until we worked out why.
Additional needs to support a student with problem behaviour in the classroom.
The environment played a large role in what Melissa was able to do and how much attention she could provide for supporting Joe:
[another student] gets loud, and that ricochets onto [another student].... and it's a bit of a chain reaction. ... So I mean all that sort of behaviour just, it just ... creates havoc in the classroom.... and because it was such a domino effect, I've never seen that happen quite so quickly and with such, with everybody being treated in some way by another child.
Melissa was certain that team agreement was an important contributing factor to supporting a student with problem behaviour:
[if the team] isn't cohesive and doesn't agree on strategies, nothing works and it just creates a lot of friction... the room doesn't work properly unless we're all on the same page and we're all working together towards the same objective
Melissa discussed the importance of having the ability to observe and reflect:
Within the hustle and bustle of the day, because it's always a very busy day, just to take that step back and do a bit more observation ... I've found that being able to ... watch that video told us a lot that we hadn't actually picked up in the chaos of the day. So it's a good time to reflect and to analyse and discuss.
When asked if Melissa would be confident in using the video's to conduct an FBA, she responded:
Yeah, probably, with a bit of practice and a bit of experience, yeah. .... Although you were good at leading us through and explaining different things....
Overall Melissa was having a hard time with the Behavior Imagingâ„¢ system and the general behaviour management of her classroom. She also identified the importance of structuring an intervention program and accessing extra support when needed:
I'm not functioning, like I normally would... I just didn't cope ... I was starting to feel really inadequate.... what I was wanting was support from other areas that I wasn't actually getting at that time. ... And so I spoke to appropriate staff ... and it was all resolved. But it's been a very tense time. ... I was just getting deeper and deeper into this feeling that this room was just so out of control. .... then I was given help, then I was assisted and then things changed. ... It's just- ... I was too proud, ... to admit that things weren't right in the room. And I tried to work it and it didn't work.
It was difficult. And I think, too, because the video and the whole scenario of the study has been very intense and something very different to ... what I've ever experienced. ... I just felt like my goodness I'm on show ... What are my teaching strategies like? I was looking ... At my whole practice of everything. I know that you said it was concentrating on the children's behaviour ... But I felt like I was on show. Like the spotlight was on me.
[when implementing a intensive strategy] we would need another staff member to shadow Joe and always be there ready to direct him if needed, so a teacher to teach him that skill. .... [throughout the year] Maybe we missed some times, some opportunities where we could've redirected him, just in the mix of the room because they're a very busy, active little group of people. It just has to be probably structured a little bit more.
The purpose of this study was to utilise mixed method approaches in comparing the validity, utility and feasibility of PBS approaches in the classroom. This included using an experimental design to compare three forms of Functional behaviour assessments (FBA) and determine the function of problem behaviour; selecting a function based intervention to reduce the student's use of problem behaviour. In addition the teacher's views of all methods were explored using a qualitative case study design to provide further understanding of the barriers and facilitators towards using PBS practices in the classroom.
The results partially support previous research findings. Quantitative results indicate that Direct Observation assisted via video Behavior Imagingâ„¢ technology provided the most accurate results of behavioural function; this is supportive of some previous findings (Alter et al., 2008; Tarbox et al., 2009). The indirect FAI was the least accurate as has been found in the literature (Alter et al., 2008; Tarbox et al., 2009), and the experimental FA provided inconclusive to no significant results, due to the behavioural function being related to a peer, who was difficult to direct as needed for the experimental manipulation trials of the FA. The function of Joe's hair pulling behaviour was identified to be maintained by the presence of an identified peer. This finding was confirmed by a direct relation between function-based environmental modification and a substantial reduction in Joe's need to pull hair in comparison to baseline. The final support strategy implemented was moving Joe to another age appropriate classroom where he did not need to rely on hair pulling behaviour as the identified peer was not present. At a two week follow up it was confirmed that Joe had not pulled hair in his new setting and was much happier coming to school.
Qualitative results illuminated information that contributed to the understanding of teacher perceptions of students' behaviour; the process of conducting an FBA in the classroom; selecting interventions; and what additional needs might be required when supporting a student with problem behaviour in the classroom. An analysis of the semi-structured interviews conducted with Melissa indicated changes across time in how Melissa viewed each process.
When understanding Melissa's perceptions of students' behaviour it became clear that initially Melissa was not sure of why Joe was pulling hair and contributed some of her understanding to his individual characteristics (e.g. he is an opportunist), after conducting all three methods of the FBA, Melissa was certain she understood why and began to describe Joe's behaviour as a result of an environmental factor (e.g. the other student is causing this). Melissa appeared to have difficulty generalising the understanding of behavioural function to other students', during the second and third interviews, she described behaviour of other students as an intrinsic characteristic 'they take advantage', or 'that's her'.
Melissa's discussions regarding the FBA process revealed some informative views. Melissa believed that the experimental FA trials were too clinical and not suited to the classroom. She found that watching the Behavior Imagingâ„¢ videos to be the most useful. She recognised that more observation of antecedents was needed, but the busy classroom, when she was teaching did not provide the opportunity for such in depth behaviour analysis. Melissa also identified that because the identified peer was unpredictable, the function of the behaviour can be masked and therefore made it difficult to identify the core function. She provided further insight in identifying that the videos were not the only step in the FBA, but that it helped in knowing what to look for during the busy day.
In regards to selecting an appropriate intervention, Melissa had grave difficulty prior to the commencement of the study. Initially she hoped the behaviour would stop on its own. She was highly reliant on consequential reactive strategies. She then proceeded to mix in some strategies with a focus to increase positive behaviour. Notably, Melissa used a repertoire of strategies that she was familiar with, but that were not individualised to Joe or were shown to be unsuccessful. Melissa admitted she was just trialling approaches. In later interviews, once the function of the behaviour was known, Melissa realised that these strategies were not appropriate. However, again Melissa had difficulties generalising this approach to other students, believing that removal and consequential strategies were the most appropriate, particularly as it was the only way to prevent students modelling problem behaviour from each other. Understanding the function of Joe's behaviour allowed Melissa to identify environmental strategies without a great deal of thought. Using Umbreit's model appeared to support Melissa in understanding that replacement behaviour was needed. Although Melissa discussed possible and suitable strategies that could be implemented, further discussion cannot be made regarding them as they were not implemented during the time of this study. The school took an entirely environmental approach to supporting Joe, he was moved to another classroom where he was happier and his access to learning and education was re-opened.
The final theme that this study was drawn to, explored the additional needs Melissa identified that were important when supporting a student displaying problem behaviour. In the interviews, Melissa brought forward some influencing variables; the classroom environment needed to be structured and calm (with little interfering attention or problem behaviours from other students); the classroom staff and team needed to be cohesive and work together when managing problem behaviour; lowered self-efficacy, reduced the teacher's ability to function as she normally would, she believed that support from principals, assistants and other teachers was crucial when she had feelings of lowered self-efficacy; and finally finding the time to observe and analyse possible behavioural function, with some possible initial guidance was also important.
Combining Quantitative and Qualitative Results
In line with previous research, Melissa did not use strategies that were individualised to Joe or the function of his behaviour (Blood & Neel, 2007; McIntosh et al., 2008; Scott et al., 2005). Melissa was initially overly reliant on consequential and removal strategies, which is consistent with the field (Scott et al., 2005; Van Acker et al., 2005; Wubbels, 2011). Melissa's lack of understanding of the behavioural function and over-reliance on consequential reactive strategies did not prove to be beneficial for reducing Joe's need to pull hair, as shown during baseline.
Three forms of FBA were used in this study, indirect, direct and experimental FA. Melissa found direct observational methods were the most useful, however only with the support of Behavior Imagingâ„¢ technology (Reischl & Oberleitner, 2009). Melissa found that the classroom was too busy to properly observe, as would be required for an FBA and has been discussed by other teachers (Emmer & Stough, 2001). The Behavior Imagingâ„¢ technology provided a means to observe at a time more convenient, that did not require too many planned changes throughout the class day, as was expected by the experimental FA (McDonald et al., 2012; Sigafoos & Saggers, 1995) . It also provided a tool to help Melissa hypothesise possible behavioural functions that she was better able to observe during the classroom day once she was aware of what she was looking for. Melissa found that she would be more inclined to use the Behavior Imagingâ„¢ methods over other methods in future if the need arose.
The results of the FBA partially support the results of previous FBA comparisons reported in the literature (Alter et al., 2008; McDonald et al., 2012; Tarbox et al., 2009). Due to the behaviour being functionally related to an identified peer, it was difficult to set up experimental conditions as required by the experimental FA and thus the results were inconclusive. It is also difficult to know if Melissa would have had a different view of the FA had the behavioural function and outcome been different.
Once the behavioural function was known, the teacher made many environmental modifications that related to the function. Although student replacement behaviours were discussed, these were not implemented. A completely environmental antecedent based approach was adopted (Stitchter, Randolph, Kay, & Gage, 2009; Wehby & Lane, 2009), which showed a dramatic reduction in Joe's need to pull hair.
In addition to the above findings, Melissa provided insight into some clear practicality issues as perceived by her. The results are not dissimilar to those communicated in other teacher perspectives research (Boardman et al., 2005; O'Neill & Stephenson, 2010) Ease of implementation, the personal feelings Melissa held, the beliefs about the appropriateness for the student and classroom as well as the availability of on the job collegial training and support all played an important role in whether she would use an approach again.
Melissa appeared to have difficulty generalising the FBA practices to other students and felt that because students had intrinsic behavioural characteristics the FBA method was not appropriate. This has ramifications for teachers that may not identify when and which students may require individualised behaviour support based on a FBA.
The results of the current study provide new considerations when conducting an FBA in a classroom environment. This study has shown the limitations of using experimental FA in the classroom not only from a practical view but from a teacher perspective. The results of conducting a direct ABC analysis using Behavior Imagingâ„¢ indicate a potentially useful tool that teachers may use in future FBA application and research, in the absence of specialist researcher support.
The results also have implications for training. The results indicate that a teacher with relatively no understanding or previous experience of FBA was able to identify the function of behaviour, and select appropriate function based interventions. However, the busy environment meant that the teacher had to find time to actively observe and analyse behaviour. The environment also played a big role in how the teacher would implement selected interventions. The teacher also identified that time out strategies were required as students might model inappropriate behaviour from one another. Future research may need to take a closer look at teachers' views of strategies, particularly exclusionary ones in relation to class wide behaviour management.
The study was not experimental in nature, it was decided that the approach would be a naturalistic study design, in which the teacher had more control over the development and implementation over her own procedures with guidance from the researcher if needed. Although the results show success, the strictly non-experimental nature of the study cannot infer direct cause and effect results with strong validity. Another limitation to the study, was in regards to the experimental FA. The FA was not appropriate due the function being related to a peer, the dislike for the FA approach expressed by the teacher may have been different had the FA provided a clearer functional understanding of the behaviour.
This project followed an intrinsic case study design. As such the results are only representative of this case; another school, teacher, or student may have yielded very different results. Therefore these results cannot be generalised and more studies are required to meet data saturation and for all alternative perspectives to be considered.
May also explain why the mixed results exist in FAI research