Functional assessment will refer to the “the full range of strategies used to identify the antecedents and consequences that control problem behavior” (Horner, 1994, p. 401).
Alternatively, the term "functional" or "experimental analysis" refers to a more specific procedure wherein environmental variables are systematically manipulated in an experimental fashion to evaluate hypothesized functional relationships (Vollmer & Northup, 1996).
Functional assessment refers to a broader set of procedures that may or may not include an experimental analysis.
For example, an experimental analysis of a child's inattentive behavior would involve the planned introduction and removal of hypothesized maintaining stimuli (e.g., teacher attention) to determine whether the frequency of the target behavior varies as a function of teacher attention.
In contrast, a functional assessment consists of a variety of techniques (e.g., interview with teacher, direct observation of the relation between target behavior and environmental events) that may or may not include direct manipulation of environmental events, designed to determine what antecedent and consequent events could be functionally related to the inattentive behavior.
From a clinical perspective, the overriding goals of a functional assessment are to determine:
(a) the function(s) of problematic behavior (i.e., what purpose is the disordered behavior serving for the individual?)
(b) environmental events or stimuli that increase the probability that problematic behavior will occur.
The term “challenging behavior” has been defined by Emerson (1995, 2001) as “culturally abnormal behavior of such an intensity, frequency or duration that the physical safety of the person or others is likely to be placed in a serious jeopardy. It can also mean the behavior which is likely to seriously limit the use of, or result in the person being denied access to ordinary community facilities”. Commonly reported challenging behaviors are aggression, property destruction and self-injurious behavior (Brosnan & Healy, 2011; Csorba, Radvanyi, & Dinya, 2011)
Many methods of functional assessment can be applied to evaluate challenging behaviors generally in subjects with development disabilities and in children with emotional and intellectual disabilities.
According to Sipes & Matson (2012), it is possible to gain information about the function of a behavior using direct methods, analogue methods and indirect methods.
The direct method has the higher ecological validity but can only use correlational data because the environment is not directly controlled. Some problem behaviors occur at low frequencies and as a consequence, they take up too much observation time. An example of direct methods are scatter plots and ABC charts. In the scatter plots, the data is collected by defining a specific interval and then obtaining information on the frequency and duration of behaviors (Touchette, 1985).
Scatter plots are simple to collect but little information can be gathered about antecedents and consequences. Psychometric properties for accurately establishing and maintaining operations of problem behavior with this approach are unknown.
In ABC charts, the observer collects data on the antecedent (A), the behavior (B) and consequences (C). According to Bijou (Bijou, Petersen, Ault, 1968) the first step of this observation process is set objective criteria to identify the possible responses of a category of behavior. Bijou underlines the importance of transforming narrative observation into systematic descriptions of the antecedents, behaviors, and consequences that occurred during the observation period. In the Bijou perspective, a behavior is based on continuous and reciprocal interactions between behavior and environmental events.
ABC charts gather a lot of information about events but low temporal information is gained, they are time consuming and subjective information can contaminate the findings.
The analogue assessment is generally referred to the Experimental Functional Analysis or EFA by Iwata (Iwata, Dorsey, Slifer, Bauman, & Richman, 1994). EFA consists in an assessment phase in a controlled environment, a random multi-element design in which the therapist manipulates several conditions (attention, escape, alone, play) and the hypothesis formulation about maintaining functions. This method has a great ability to infer causation but it requires the occurrence of a behavior, it could stimulate the emergence of new functions of the behavior and it is not easily practicable in applied settings.
The experimental functional analysis is very time-consuming, requiring lengthy periods of time to be completed (generally between 40 and 60 sessions). Brief functional analysis has been developed, adapting the classic functional analysis procedure. The brief functional analysis consists of a 90-min session in which researchers use a multi-element design across potential maintaining conditions followed by a replication phase of assessment.
In the most of cases, the brief functional analysis allows researchers to identify the maintaining function of a behavior. However this method seems to be limited to those patients who show high-frequency behavior (Derby, Wacker, & Sasso, 1992).
Another analogue method is the concurrent choice methodology (Harding, 1999). It is used to perform a functional analysis evaluating the variables that sustain appropriate rather than aberrant behavior. This methodology is a potential solution to the problems emerging in the classical functional analysis procedures, such as the evaluation of low frequency behaviors or the risk to stimulate the problem behavior for the assessment. Finkel (2005) examined how choice could be used to identify behavioral function when functional analysis results were inconclusive. Using this procedure, authors were able to identify a maintaining condition for the participant’s choice behavior, thus identifying a reinforcing approach for use during treatment.
Indirect methods do not require direct observation as information is gathered through third parties such as parents and teachers. Being less intrusive, this method is more usable over a long period of time and with many behaviors. But the cause of the function cannot be determined accurately and only correlational information can be obtained.
Example of indirect methods tools are:
The Functional Assessment Interview(FAI) (O’Neill et al, 1997) allows clinicians to gather information about the topography of the behavior, setting events, other events surrounding the behaviors (i.e. sleep cycles, eating routines), the child abilities, difficulties, communication skills and previous treatment attempts, etc. The FAI requires approximately 45–90 min to administer.
The psychometric properties of the FAI has still to be studied.
The Questions About Behavioral Function (QABF) (Matson & Vollmer, 1995) is a standardized test to make functional assessment in real world settings. The scale consists of 25 items distributed in 5 factors: attention, escape, non-social, physical and tangible. Each factor is a possible function underlining a target behavior. Many studies demonstrate the strong psychometrics properties of the QABF (a recent review has been recently made by Matson, Turek & Rieske, 2012). It is good to excellent reliability, considering convergent and discriminating validity when compared to other similar scales or to experimental functional analysis. The scale is able to identify the most common functions associated to various challenging behaviors.
The Motivation Assessment Scale (MAS) (Durand & Crimmins, 1988), a rating scale designed to assess the relative influence of social attention, tangibles, escape, and sensory consequences on self-injury but it is applicable more in general for challenging behaviors. MAS is composed by 16 items which are rated by the informant (parents or teachers) on a seven-point Likert scale. The MAS takes approximately 5-10 min to administer. Studies evaluating the inter-rater reliability of the MAS have found mixed results. The MAS demonstrates good test–retest reliability, good internal consistency but a low convergent validity of the MAS with EFA. Finally, there isn’t a general consensus of research about the factorial structure of the MAS items.
The Functional Assessment for Multiple Causality (FACT) (Matson et al., 2003) is a 35-item measure that was constructed to identify a hierarchy of behavioral functions for individuals with Intellectual Disabilities who present maladaptive behaviors that serve multiple functions. The FACT use forced-choice questions in order to identify and measure the strength of each behavior function. In the original study (of Matson et al ibidem) emerged a five-factor structure of the FACT: Tangible, Physical, Attention, Escape, and Nonsocial. Besides, internal consistency of the subscales and reliability for the FACT were good to excellent. There are not many other studies about the psychometric properties of FACT. They have to be still extensively studied.
The Functional Analysis Screening Tool (FAST) (Iwata et al., 2013) is used to gather from information from verbal reports about specific conditions under which the problem behavior can occur. It categorizes problem behaviors in 4 functional categories: social-positive reinforcement; social-negative reinforcement; automatic-positive reinforcement; automatic-negative reinforcement.
The FAST consists of 3 sections. The first collects information about the client, problem-behavior and the client-informant relationship. The second provides 16 dichotomous questions about antecedents and consequences of the problem behavior. The final section provides a scoring summary.
According to Iwata, reliability of the FAST is moderate at best as well as its validity, considering that the correspondence between FAST and functional analysis was 63.8%.
One of the most useful techniques to gather information about pupil’s behavior, both generally and specifically, with reference to identified behaviors is through direct observation. This can be used to verify the pupils’ targeted behaviors and to gather further information about other behaviors that they display. This provides the ability to compare them with their classroom peer group.
The initial step during a direct observation is to gather clear and concise descriptions of the behaviors being observed. This clarity will lead to a better understanding between the professionals and caretakers. This also promotes clarity while collecting the data as all involved are fully aware of the specific information that they are collecting. This will help to eliminate problems that might arise between ‘where and when’ the observations were taken. Haynes and O’Brien (2000) provided guidelines for defining the different target behaviors (dangerous, restrictive, interfering, etc.).
It is important to identify which parts of the pupil’s behaviors are significant and need to be focused on. This will help maintain the integrity of the observations and the data collected. Once this has been determined, the behavior must be operationally defined with sufficient detail that observers reliability can be obtained (Hurwitz & Minshaw, 2012).
While we may have identified the behaviours to record the observers will need to judge if a particular behavior is of sufficient importance to warrant being recorded. Alongside of this is the need to decide if they should consider other significant aspects that may be impacting on the pupil being observed. These behaviors may be socially or when learning. Both parents and teachers play a significant part as they can help identify the significant behaviors and the drives, which may be powering those behaviors. They also have a good knowledge of the pupil’s social world and how personal relationships may affect them.
Issues in Behavioral Observations
No matter what behavior recording system you decide to use you will gather useful information about your pupil’s behavior in school and the classroom and during times that they are relating to others. Initially, it’s useful to ask the teachers to identify specific times and places that the ADHD behaviors appear and how long they may be present for. It is useful to observe pupils during challenging times when the pupil has to attempt to suppress the ADHD behaviors such as impulsivity or hyperactivity for example. Often this is during the more formal lesson times or during extended periods of time. It is also helpful to observe during more relaxed free flowing lessons to obtain a contrast of the pupil behaviour. During the free flowing lessons ADHD pupils behaviors will resemble their non-ADHD peers. The observations in less formal observations will not be helpful in collecting information about their ADHD behaviors but will serve as a useful contrast.
It is important to understand the social world the ADHD pupil is working and learning in. This will provide a ‘benchmark’, which will help illustrate how the identified pupil is moving away from the expected peer group behavior. An example of how this could be used would be to observe one of the peers who have been identified as having the average type of behavior and the ADHD pupil. Time is spent observing one pupil then a comparable amount of time observing the identified pupil. The pupil being compared to the ADHD pupil is changed so that the observer is randomly following different children during his observations while continuing to return to the ADHD pupil. This will help create more of a sense of the class average.
There are no firm guidelines laid down for how long and how often to observe a pupil during a ADHD observation. This is probably best agreed during the initial teacher consultation. As a general rule I would expect to observe on two or three different school days for between 20 to 30 minutes. However it may be
necessary to conduct specific observations during specific lessons. Small group work may be show that they are impulsive and chatty but the math’s lesson in the classroom may present specific concentration issues that lead to a different behavior. It may be necessary to observe these classes on a couple of occasions.
After the observation it is helpful if the observer can write a short overview of the day. Commenting on the pupils motivation, teachers approach, pupils interest in lessons, relationship with peer group, any unusual disturbances (classroom visitors), etc. Also make a note the classroom atmosphere. Is it hot or cold, has the lesson been well prepared, is there any evidence of accommodations being made for pupils that need additional help.
Observation methods
Once the behaviors to be observed and monitored have been identified we will need to choose the best and most appropriate method of collecting and recording the data. Care must be taken to record and observe the behavior we really want to observe and not be sidetracked away from our targeted behaviors.
Wright (2002) recommends that any ADHD observation should follow the three prime behaviors isolated by Platzman et al. (1992). These are excessive motor activity, the pupil’s negative vocalizations, and any off task behaviors noticed.
Wright in his online manual (ibidem) describe the different methods of recording as following:
Event or frequency recording. This allows the observer to build up an accurate picture of the frequency of the identified behaviors provided the behaviors have clear beginnings and ends (i.e. throwing things in class, out of seat behavior, a single vocalization). The time of the observations can be as short as ten minutes or be part of a whole day observation.
Interval recording. Through this method is possible estimate the duration or length of a behavior. Observers have to specify time periods (intervals) and then record on a sheet the presence or absence of behavior during all the period of observation. It includes three type of recording methods following described.
Generally, time intervals of 10 to 15 seconds are suitable to capture information on behaviors.
Whole interval recording. The examiner observes and notes when the behavior occurs throughout the entire interval. This means the observer only marks a behavior as happening IF it is present throughout the period of the observation time. An advantage of this type of observation is that it provides an idea of the time passing while the behavior is in place. However it fails to record the behaviors that do not persist throughout the identified period of observation. This type of observation is perhaps more effective when observing positive behaviors such as time at work after a teacher intervention.
Momentary time-sampling. This method is a subset of the previous and it can be helpful when observing behaviors that do not have a specific identified onset or end. Once the length of the interval is settled, the observer’s task is record the occurrence or not of a behavior at the very beginning or at very end of each time interval. They then record the result on a data sheet. This requires the observer to look at the pupil and record the behavior they notice at that specific time. This has the advantage of being accurate at the time of observation but will inevitably lead to missing other behaviors during the non- observed time, when recording.
Partial interval recording. This procedure aims to record if the behavior occurred at least once during the short observation interval. This type of observation has the advantage of collecting data on the observed behaviors when they happen. This is useful for noting if there is a sudden change in negative (or positive) behaviours after a specific event. This type of observation can lead to an impression of overestimation of the more negative behaviors. This is often because observers sometimes are intent on catching every specific occurrence of a behavior. Often this is used to collect information on the number of times a negative behavior, such as aggression occurs. There may not be a differentiation in the level of the aggression.
After choosing the behaviors to be observed and methods of observing the observer decides on the length of time of the observation. There are advantages to short time intervals as it allows the finer path of the behaviors to be observed. The downside though is that the observations are reduced to say 10 to 15 seconds a series of quickly written symbols are collected that then need to be interpreted. The longer observational periods allow for more information to be collected and written in hard copy. But this allows some behaviors to be missed during the writing of the observed period and that this lost information could play a useful part in an ADHD observation.