Keywords

FormalPara Topics Covered Within This Chapter

Topics

Functional Analysis Procedure

Interpreting Functional Analysis Results

Advantages and Limitations of Functional Analysis

The purpose of a functional behavior assessment is to identify the reinforcement maintaining challenging behavior so that this information can be used to develop effective, individualized interventions. The identification of the specific consequences positively or negatively reinforcing target challenging behavior will allow supervisees to design interventions to teach alternative, socially acceptable behaviors to access the same reinforcers. Additionally, this allows supervisees to minimize access to reinforcing consequences upon the future occurrences of the challenging behavior. As a result, your supervisees will need to be proficient in implementing and interpreting a number of functional behavior assessment methodologies, including indirect assessments, direct assessments, and experimental functional analyses. While indirect and direct assessments allow clinicians to develop hypotheses regarding the environmental events maintaining challenging behavior, the functional analysis is the only approach to experimentally evaluate those hypotheses. The functional analysis is the most reliable and valid functional behavior assessment methodology. Needless to say, functional analyses have become the bread and butter among behavior analysts because of their utility.

Iwata et al. (1994) first developed this approach to identify the functional properties of self-injurious behavior. Since then, functional analyses have become a staple assessment among behavior analysts, used to assess a myriad of topographies of challenging behavior. A functional analysis is designed to mimic the naturally occurring environmental variables that may occasion for and reinforce challenging behavior. Functional analyses provide the opportunity to systematically manipulate environmental events similar to those in the client’s natural environment in order to determine if they have functional relation with the target challenging behavior.

Functional Analysis Procedure

At a minimum, a functional analysis contains at least one test and one control condition in order to compare challenging behavior under each condition. However, functional analyses can be conducted with multiple test conditions and control conditions.

Conditions are comprised of three components: motivating operations, discriminative stimuli, and consequences. First, functional analysis conditions manipulate motivating operations so the reinforcer effectiveness of the stimuli associated with the condition are elevated. Second, each condition contains stimuli that signal the availability of reinforcement associated with that condition. Finally, each condition is associated with a specific, potentially reinforcing, consequence delivered on a dense schedule of reinforcement contingent upon instances of challenging behavior. Supervisees who grasp the basic framework of a test condition will have the skills to develop conditions that are tailored to the unique environmental events hypothesized to have a functional relation with challenging behavior among their clients.

Iwata et al. (1994) conducted three test conditions and a control condition. Today, the most common test conditions include attention, escape, tangible, and alone or ignore; yet, the possibilities for functional analysis conditions are endless, and many studies have included other conditions (Beavers et al., 2013; Rispoli et al., 2014; Van Camp et al., 2000). See Chap. 10 for a further review of additional conditions that have been included in functional analyses.

The attention condition begins with the implementer interacting with the client for a short period of time and then diverting their attention to another activity, such as reading a book. It is common to make low preference toys or leisure activities available during the attention condition. Contingent upon challenging behavior, the implementer delivers attention. As with all conditions, implementers should mimic the individual’s natural environment; in other words, deliver attention in a manner that is similar to how those in the client’s natural environment would deliver attention contingent upon challenging behavior. For example, if the client’s caregivers report responding to challenging behavior with redirecting statements use similar redirecting statements in the attention condition. On the other hand, if the client’s parents deliver back rubs to calm the client engaging in challenging behavior, back rubs should be delivered within the attention condition. Different forms of attention may affect the outcomes of a functional analysis; therefore, developing an individualized methodology, particularly in regard to the topography of attention, is important (e.g., Fisher et al., 1996).

The escape condition begins with the implementer instructing the client to complete nonpreferred tasks, such as academic tasks, and then delivering prompts in a least-to-most prompting hierarchy. Contingent upon challenging behavior, the implementer removes the demands, typically by both removing materials and orienting their body away from the client for a short duration of time. The selection of tasks is a critical step in developing an escape condition because insufficiently evocative demands could lead to a false negative conclusion. Supervisees should conduct demand assessments (see Chap. 8).

Prior to the start of a tangible session, the implementer gives the client access to a highly preferred item, such as a toy. When the session begins, the implementer removes the item out of reach, but keeps the item in the client’s line of sight, when possible. Contingent upon challenging behavior, the implementer returns the item for a short duration. The selection of the item is a critical feature of this condition. If the implementer selects a low preferred item, it is likely that its removal will not evoke challenging behavior to the same extent as the removal high preferred item, thus producing a false negative result. As a result, supervisees should implement preference assessments (Avery & Akers 2021) (see Chap. 8).

In addition to developing specific condition protocols, supervisees must decide (a) how many topographies to evaluate in a single functional analysis, (b) the duration of sessions, (c) number of sessions per condition to conduct, (d) how to measure challenging behavior, and (e) the experimental design. Individuals with intellectual and developmental disabilities who engage in challenging behavior may engage in multiple topographies (Derby et al., 1994; Derby et al., 2000). In deciding to test multiple versus a single topography during a functional analysis, clinicians must balance the need for efficiency with the possibility of invalid results. In many cases, individual topographies of challenging behavior are maintained by unique consequences; therefore, evaluating multiple topographies in a single functional analysis may preclude the clinician from accurately identifying the functions of each topography of challenging behavior.

Session duration may vary, but most commonly, functional analysis sessions are 5–15 minutes (Beavers et al., 2013). The benefit of a shorter session duration is an overall decreased assessment duration. This amounts to less time in which challenging behavior may be both evoked and reinforced. Moreover, shorter assessment duration, if valid, will lead to faster access to function-based treatment. However, the decreased session duration may result in poorly discriminated contingencies across conditions; thus, ultimately delaying the ability to develop a valid function-based treatment. In many cases, it may be best to guide supervisees to first use a shorter duration and increase the duration only if a discriminated pattern of responding among the conditions does not occur.

Needless to say, the total duration of the assessment is not only influenced by the duration of sessions, but also the number of sessions conducted per condition. Functional analyses in which two of fewer exposures to each condition are considered brief functional analyses and will be discussed in Chap. 10 (Northup et al., 1991). Supervisees conducting full functional analyses should conduct at least three exposures to each condition, but should continue administering additional sessions of a condition until a pattern of responding can be identified.

Supervisees may choose either continuous or discontinuous data collection during a functional analysis, as both approaches are well established in the literature (Beavers et al., 2013). The most commonly used continuous measurement is a frequency measure, often converted to a rate, while the most commonly used discontinuous measure is partial interval recording. See Chap. 5 for further discussion on the benefits of various approaches to measurement.

Finally, supervisees must identify an experimental design to employ within the functional analysis. A multielement design is by far the most widely used design, but research has supported the use of additional designs such as an ABAB design or a combination of designs (Beavers et al., 2013).

Interpreting Functional Analysis Results

Supervisees must not only know how to conduct the functional analysis, but also how to interpret the results. As with all single-case experimental designs, supervisees must hone visual inspection skills. Please refer to Chap. 6 for a more thorough discussion of visual inspection. In order to analyze functional analysis results, the supervisee must identify the conditions in which challenging behavior occurred at consistently higher levels in comparison to the control condition. It is our experience that supervises frequently and erroneously identify the condition with the highest level of challenging behavior as the maintaining function without considering additional conditions in which challenging behavior is elevated relative to the control condition, but not elevated as high as another condition. Conditions associated with higher levels of challenging behavior relative to the control condition suggest that variables associated with that test condition maintain that challenging behavior. It is possible for a challenging behavior to serve two or more functions. In such cases, challenging behavior will be elevated in multiple conditions relative to the control. Automatically maintained challenging behavior may present two unique patterns. First, challenging behavior occurs most frequently in the alone or ignore condition relative to all other conditions. Second, challenging behavior remains high across conditions, suggesting that variables associated with each condition have little effect on the challenging behavior. Finally, in some instances, the results of the functional analysis may be undifferentiated. That is, levels of challenging behavior are variable across conditions. This may occur because challenging behavior failed to come under control of the discriminative stimuli within the conditions and perhaps more salient stimuli would resolve this issue. It is also possible that motivating operations were not sufficiently manipulated so that the stimuli associated with the condition temporarily failed to serve as reinforcers during the analysis; therefore, motivating operations could be more appropriately manipulated. Finally, it is possible that the unique reinforcers maintaining challenging behavior were simply not presented in the functional analysis (see Carr et al., 1996; Rispoli et al., 2014 for examples). As supervisees become proficient at implementing functional analyses, it is important for them also to develop skills in troubleshooting undifferentiated functional analysis results.

Advantages and Limitations of Functional Analysis

Supervisees must also recognize the advantages and limitations of functional analyses in order to make sound judgments about when to conduct a functional analysis. The most obvious advantage of a functional analysis is that it produces valid conclusions regarding the function of challenging behavior, allowing clinicians to develop function-based interventions. However, there are disadvantages that supervisees must also recognize and consider. First, functional analyses require resources, particularly time and effort of highly skilled behavior analysts. There may be situations in which supervisees should allocate time and resources that would be needed to conduct a functional analysis to other activities to best serve the client. Additionally, it is possible that a functional analysis may, temporarily, strengthen challenging behavior. If the challenging behavior persists in the client’s natural environment, it is clear that the challenging behavior is being reinforced in some capacity outside of the functional analysis; however, it is important to recognize the possibility of strengthening a maladaptive behavior. Preliminary research suggests that in many cases, a functional analysis does not affect levels challenging behavior outside of the functional analysis setting (Davis et al., 2014), but further research is warranted. Finally, within a functional analysis, we expect to observe high levels of challenging behavior, which increases the risk of injury to the client and implementers. Appropriate measures should be taken to mitigate this risk as much as possible. For a more thorough summary of risk mitigation practices, see safety measures described in the following resources and studies: Iwata et al. (1994), Kahng et al. (2015), Lalli et al. (1995), Marcus et al. (2001), Matson (2012), Poling et al. (2012), Wallace et al. (1999), Weeden et al. (2010), and Wiskirchen et al. (2017).

Group Supervision Meeting

Below is a plan for activities to incorporate into a 1-hour meeting with a small group of supervisees.

Group Supervision Meeting Agenda

Time

Activity

0:00–20:00

Review Major Concepts

20:00–40:00

Develop Functional Analysis Protocol

40:00–55:00

Interpret Functional Analysis Results Graphs

55:00–60:00

Knowledge Check

figure a

Reading Assignments

At least one week prior to the group supervision meeting, assign your supervisees to read about the subject. Below is a list of recommended assigned readings.

  • Beavers et al. (2013)

  • Hagopian et al. (1997)

  • Iwata et al. (1994)

Review Major Concepts

Begin your group supervision meeting by reviewing major concepts associated with conducting and interpreting functional analyses. Begin the conversation with identifying what constitutes a functional analysis and reasons for conducting a functional analysis. Next discuss the components of a test and a control condition. Finally, lead a discussion on how to make decisions regarding: (a) number of topographies to assess in a single functional analysis, (b) data collection method, (c) session duration, (d) number of sessions per condition, and (e) experimental design. Below is a summary to guide those discussions and PowerPoint slides are available to share with your supervisees.

Ask supervisees to define a functional analysis. Within their definition, ensure that they identify the following features of a functional analysis:

  • It is an experimental evaluation of how specific variables influence challenging behavior.

  • It is the most sophisticated, reliable, and valid functional behavior assessment tool.

  • It allows clinicians to identify antecedents and consequences associated with a target behavior.

  • The results of a functional analysis can inform a function-based intervention.

Ask supervisees to discuss why they may wish to conduct a functional analysis. Be sure that if supervisees plan to implement a function-based intervention, they must first conduct a functional analysis. This conversation may also include a discussion of specific clients because functional analysis considerations are made on a case-by-case basis. Various case-specific considerations may be discussed such as topography of challenging behavior, context in which challenging behavior occurs, the ability to mitigate risk to the client and implementer, the supervisee’s experience and expertise, caregiver or stakeholder preference, and much more.

It is critical that supervisees develop a deep understanding of how a functional analysis operates. Encourage your supervisees to avoid the urge to simply read and memorize a protocol provided in a journal article or by an employer. Instead, guide your supervisees to grasp the logic behind the procedures. To facilitate this understanding, use the Developing Functional Analysis Test and Control Conditions (Appendix A) to identify the three major components (establishing operation, discriminative stimuli, and reinforcer) of the four most frequently implemented test conditions: attention, escape, tangible, and ignore, as well as control condition(s). It is best to begin with the test portion; for example, (a) asking your supervisees to identify how they could manipulate the environment in order to facilitate an establishing operation for attention, (b) what stimuli would signal the availability of attention, and (c) how to deliver attention as a reinforcer. After identifying the main components of the test condition, it is easier to develop the control condition because this is essentially the opposite of the test condition. After completing all four conditions, discuss the fact that a traditional functional analysis (Iwata et al., 1994) consists of a single control condition that combines the abolishing operations, SΔ, and extinction procedures associated with all three control conditions they developed. Engage your supervisees in a discussion of the benefit of the combined single control condition as opposed to three unique control conditions.

Finally, discuss how to make decisions regarding: (a) number of topographies to assess in a single functional analysis, (b) session duration, (c) number of sessions per condition, (d) data collection method, and (e) experimental design. In discussing each of the five items, communicate to supervisees that clear-cut guidelines simply do not exist. Rather, they must use their knowledge of functional analysis combined with their knowledge of the needs and wishes of client and the client’s caregivers to identify the best methods for each individual case. In other words, 5-minute session duration may be the best choice for one client, while designing 15-minute sessions is best suited for another client. Therefore, this discussion should be focused on factors to consider rather than rules to follow.

In determining the number of topographies to assess in a single functional analysis, the supervisees should consider the likelihood of inconclusive results if multiple topographies are to be assessed simultaneously. If two or more topographies serve two or more functions, but are assessed simultaneously, it is likely that the results will not accurately identify the function for each individual topography; rather, it is likely that the results will appear to be inconclusive. In some cases, anecdotal evidence gathered through indirect and nonexperimental direct functional behavior assessments indicate a likelihood of a similar function among two or more topographies, thus providing a stronger case for a analyzing the topographies simultaneously. Nonetheless, we recommend that you encourage your supervisees to limit functional analyses to a single topography until they gain more experience. After gaining more experience, they may be able to successfully identify cases in which a functional analysis on multiple topographies could prove successful, but it is unlikely that they would be able to make this distinction early in their careers.

As with other procedural decisions, session duration decisions should be made on a case-by-case basis. You should guide supervisees to identify factors to consider. These factors include, but are not limited to, the client’s prior history in discriminating contingencies and frequency of challenging behavior. Clients who need longer exposure to facilitate discriminated responding will necessitate longer session duration. This may also include the frequency of challenging behavior, with more frequently occurring challenging behavior being better suited for shorter sessions and less frequent challenging behavior necessitating longer sessions. You may discuss other factors that that you as well. A good rule of thumb to communicate to supervisees is, unless otherwise indicated, to begin with a shorter duration and only increase if discriminated responding failed to occur or if few or no instances of challenging behavior occurred.

The decisions related to selecting a data collection method gives the supervisor the opportunity to review content introduced in Chap. 5. Use this opportunity to assess for maintenance of those skills and reteach as necessary. Similarly, determination of experimental design selection and how many sessions to conduct per condition provides an excellent opportunity for the supervisor to review concepts of experimental design and visual analysis introduced in Chap. 6.

Develop a Functional Analysis Protocol

Assign your supervisees to groups of three or four. Instruct the groups to create a protocol for a traditional functional analysis that will contain the following conditions: (a) attention, (b) tangible, (c) escape, (d) ignore, and (e) a control condition. They should create their protocols for each condition on the Functional Analysis Protocol and Procedural Fidelity Checklist (Appendix B). We recommend that they create their protocols electronically so that they can easily share the completed document with for use an upcoming role-play.

Allow your supervisees at least 15 minutes to develop their protocol, but provide them with a 3-minute incremental signal so that they can devote the same amount of time to each condition. Once they have finished their protocols, use the last 5 minutes to review them carefully. We recommend you require each team to send you an electronic version as they complete them so you can begin to edit one condition at a time. They will use these protocols for their next role-play activity; therefore, any errors need to be identified before they are practiced. Once each small group has a supervisor-approved set of functional analysis protocols, be sure each of the group has a copy. Please note that if your supervisees are in a field experience placement that uses a specific functional analysis protocol, your supervisees can use that protocol, rather than the one developed in the group, during the upcoming role-play.

Interpret Functional Analysis GraphsDistribute the Interpreting Functional Analysis Graphs (Appendix C). The graphs are also incorporated into the PowerPoint slides for ease. For the first eight graphs, vocally describe to your supervisees how to interpret the graphs. For the last seven graphs, have them interpret the graphs independently. When supervisees indicate they completed their analysis, ask them to share with the group. We recommend the following steps to interpreting the functional analysis graphs, which are loosely based on procedures outlined by Hagopian et al. (1997) and Roane et al. (2013). However, you may consider adopting the exact procedures outlined by these authors. The rigorous procedures outlined by Hagopian et al. and Roane et al. are most likely to produce accurate conclusions. However, for most clearly differentiated results, the following procedures are likely to produce accurate results, which, in addition to their ease of implementation, make them appropriate for most functional analysis interpretations. Therefore, we recommend utilizing the procedures described in Table 9.1 for interpreting most functional analysis graphs and adopting the Hagopian et al. and/or Roane et al. procedures with less clearly differentiated results and any other situation in which you feel this will improve accuracy of functional analysis result interpretation.

Table 9.1 Steps to interpreting functional analysis results
Full size table
figure b
figure c

Individual Supervision Meeting Without a Client

Below is a plan for activities to incorporate into a 1-hour meeting with an individual supervisee.

Individual Supervision Meeting Without a Client Agenda

Time

Activity

0:00–10:00

Review Functional Analysis Graph and Interpretation Homework

10:00–60:00

Role-Play Functional Analysis and Performance Feedback

figure d

Review Functional Analysis Graph and Interpretation Homework

Review your supervisee’s graphs using the Graph Component Checklist (Appendix E). It is helpful for your supervisee to watch you analyze the graph according to this checklist. As you do so, provide specific praise to your supervisee for inclusion of the components and provide both specific feedback and rationale for any components that are not incorporated into the graph. After checking for the components of each graph, ask your supervisee to share their interpretation of the results (i.e., function maintaining challenging behavior). Again, deliver specific praise for correct interpretations. In response to errors, model for the supervisee how to determine the correct function of challenging behavior, providing a rationale for each step in the interpretation process. Correct interpretations of the functional analysis data sets are found in Table 9.2.

Table 9.2 Functional analysis interpretation activity: Correct interpretation
Full size table

Role-Play Functional Analysis and Performance Feedback

The final portion of your individual supervision meeting will involve a role-play of all functional analysis conditions. Role-play each condition in the following order: (a) attention, (b) escape, (c) tangible, (d) ignore, and (e) control. Prior to beginning the role-play, ask your supervisee to provide you with the protocols, if they have not done so already. Quickly read the protocols aloud together to ensure that you both have the same protocols and that all clarifications have been made prior to the role-play activity.

You will role-play as a client and your supervisee will implement a functional analysis. We suggest you select a challenging behavior topography that the supervisee is likely to encounter at their field experience placement as opposed to a confederate topography (e.g., tapping the table) that is not authentically challenging. That being said, the intensity of the behavior should be modified so there is no risk of injury to either you or the supervise. For example, touching one’s nose should not be considered as a challenging behavior for the role-play because this is not an authentic topography of challenging behavior. On the contrary, you may choose self-injury as the target challenging behavior due to its authenticity. However, if you selected head hitting as the target topography during role-play, engage in head hitting with an open hand and no force so that in reality you are only tapping your head. We also suggest that you maintain a moderate frequency of target challenging behavior and only occasionally engage in nontarget challenging behavior. You want to role-play a realistic functional analysis, but one that is relatively easy to implement. You want your supervisee to be successful during this initial role-play. As you observe your supervisee’s success, you can begin to emit behaviors that would increase implementation difficulty. Some examples of ways to increase difficulty of accurate implementation include: (a) increasing the frequency of the target behavior, (b) engaging in nontarget challenging behavior, (c) continuing to engage in challenging behavior when accessing the condition-specific consequence (e.g., continuing to hit your own head even when you have access to the tangible item), and (d) other scenarios that have threated your own fidelity of implementation that you have experienced as a clinician.

Your supervisee will role-play as the implementer. During the role-play, do your best to refrain from coaching your supervisee. Instead, hold your comments to the end of the condition. Of course, use your best judgment regarding when to support a faltering supervisee mid-role-play. Ideally, you would measure procedural fidelity as you are simultaneously playing the role of the client. In reality, measuring procedural fidelity while simultaneously role-playing the client may be distracting or impossible. Therefore, if it is impossible to record procedural fidelity live during the role-play, simply review the procedural fidelity checklist with your supervisee immediately after the session. For the sake of efficiency, we recommend engaging in a 5-minute role-play condition, followed by 5 minutes of feedback, utilizing the procedural fidelity checklist. With five conditions, each requiring 10 minutes for role-play and subsequent feedback, this entire activity should take about 50 minutes.

At the end of the role-play, be sure to emphasize again the steps your supervisee completed correctly. Also review the steps the supervisee needs to practice again. If procedural fidelity was below 80% of steps correct for any condition, we recommend that you ask your supervisee to schedule a second role-play opportunity for those condition(s). Encourage your supervisee to role-play with peers in preparation for this second opportunity. We highly encourage you not to allow any supervisee to implement a functional analysis with clients until the supervisee can complete at least 80% of steps correctly across all conditions, regardless of the amount support that will be in place for your supervisee’s first attempt in implementing a functional analysis.

figure e

Individual Supervision Meeting with a Client

Below is a plan for activities to incorporate into a 50-minute supervision session in which you observe your supervisee with a client.

Individual Supervision Meeting with a Client Agenda

Time

Activity

0:00–5:00

Review Procedural Fidelity Checklist and Data Collection Procedure

10:00–50:00

Conduct Functional Analysis with Support

figure f

Review Procedural Fidelity Checklist and Data Collection Procedure

During this meeting, you will support your supervisee in implementing a functional analysis. Prior to conducting the functional analysis, review the procedural fidelity checklists and data collection procedures with your supervisee. You can conduct this review any time prior to conducting the functional analysis, not necessarily immediately prior to the functional analysis.

Conduct Functional Analysis with Support

It is quite possible that your supervisee is not currently working with a client in need of a functional analysis at the exact timing of this supervision session. If that is the case, we encourage you to solve this with one of two options: (a) determine if a client at your supervisee’s field experience placement is in need of a functional analysis and if your supervisee can volunteer to assist in implementation, or (b) role-play a functional analysis that is more authentic than the previous supervision meeting with the assistance of additional supervisees. We highly recommend the first option because it is a more authentic learning experience. Additionally, the second option precludes the ability for this session to take place with a client; therefore, additional supervision meetings in which you observe your supervisee working with a client will be necessary. If you choose the second option, conduct the role-play with three supervisees. One will implement the functional analysis, one will play the role of a client, and the other will collect data. As a supervisor, you will collect data to evaluate IOA with the data collector as well as complete procedural fidelity checklists during implementation. You may increase the difficulty of implementation relative to the prior supervision session with strategies discussed in the previous section.

Ideally, your supervisee will implement a functional analysis with a client. During this implementation, client safety and accurate implementation are key. Therefore, include individuals with sufficient experience to implement the functional analysis. Do not consider the supervisee as an implementer in determining staffing ratio because the supervisee is still in a learning phase and likely to make mistakes and need support. In other words, if an evaluation team previously determined that two implementers would need to be present during the functional analysis, then maintain two implementers and the supervisee would serve as the third. You may wish to serve as one of the implementers or only observe the implementation. This decision is yours to make as to your comfort level in both simultaneously implementing and providing feedback should be considered.

Ask the supervisee to implement functional analysis procedures for at least 25 minutes. You may want to extend the length of observation, particularly if the functional analysis consists of 10- or 15-minute sessions. During this session, ensure that your supervisee serves as the lead implementer. Do not expect that your supervisee could implement as well as collect data or self-evaluate procedural fidelity. As a new skill, the supervisee should only focus on correct implementation. During this time, encourage all implementers, including yourself, to provide immediate feedback to the supervisee, when appropriate. Errors should be corrected immediately during the session. This both ensures the validity of the functional analysis results and prevents the supervisee from practicing errors.

At the end of the observation, ask your supervisee to leave the assessment setting and visit briefly with you about the implementation. Share procedural fidelity data you gathered and specifically praise steps your supervisee implemented correctly and provide a detailed description of errors committed during implementation. When discussing errors, provide suggestions for improving implementation and a rationale as to why such improvements need to be made.

Mastery Criteria

In order to progress from this lesson, your supervisee must conduct a functional analysis in which they (a) accurately collect data with at least 80% agreement and (b) conduct the all conditions of the functional analysis with at least 80% fidelity. If either of these are not met, a second individual meeting without a client with intensive role-play and feedback should be scheduled.

figure g