Behavior observations for linking assessment to treatment for selective mutism.
Article Type:
Mutism (Care and treatment)
Mutism (Diagnosis)
Shriver, Mark D.
Segool, Natasha
Gortmaker, Valerie
Pub Date:
Name: Education & Treatment of Children Publisher: West Virginia University Press, University of West Virginia Audience: Professional Format: Magazine/Journal Subject: Education; Family and marriage; Social sciences Copyright: COPYRIGHT 2011 West Virginia University Press, University of West Virginia ISSN: 0748-8491
Date: August, 2011 Source Volume: 34 Source Issue: 3
Geographic Scope: United States Geographic Code: 1USA United States

Accession Number:
Full Text:
Selective mutism is a childhood disorder that most school psychologists and educational providers will come across at least once in their careers. Selective mutism is associated with significant impairment in educational settings where speaking is necessary for academic and social skill development. Effective treatments for selective mutism typically involve shaping or stimulus fading procedures. Choosing an effective treatment strategy for a child with selective mutism is dependent upon careful analysis of data gathered during the assessment process. This article focuses on behavior observations as a primary source of data for effective decision making regarding treatment for selective mutism. Previous literature on behavior observation and selective mutism is reviewed and guidelines are presented for decision making based on observational data. This article presents two case studies that illustrate the use of observational data for treatment decision making. In addition, the role of behavioral observations to inform selective mutism treatment decisions in practice and the need for future research on this topic are discussed.

Selective mutism is a disorder that is typically noticed during childhood and characterized by a child's almost complete lack of speaking in certain situations or settings but adequate speaking in other situations and settings, most often at home and/or with family members. The classroom is often the setting where a child's lack of speech becomes most noticeable and may be most problematic due to academic and social expectations in school. In mental health settings, the prevalence for selective mutism is reportedly low, with a prevalence of less than 1% (American Psychiatric Association, 2000; Elizur & Perednik, 2003). It is likely that the observed prevalence of selective mutism in schools may be higher, however, as the classroom is the setting in which impairment may be most likely observed (Bergman, Piacentini, & McKracken, 2002; Kumpulainen, Rasanen, Raaska, & Somppi, 1998). It seems likely that psychologists working in schools will come across at least one child with selective mutism sometime in their careers.

Although etiology is unclear contemporary research on selective mutism suggests that it is an anxiety-related disorder, or at least frequently co-occurs with anxiety (Black & Uhde, 1995; Vecchio & Kearney, 2005). Conceptualizing selective mutism as an anxiety disorder implies that there is an adverse physiological arousal of the autonomic nervous system in response to certain stimuli. For children with selective mutism, it may be that anxiety (i.e., the averse physiological response) occurs in the presence of people outside the family context. It is not clear what the controlling variable(s) is that elicits anxiety for children with selective mutism. For example, anxiety may be elicited by the mere presence of other people, or when the attention of others is directed toward the child (i.e., looking at the child, speaking to the child) or by the act of actually talking to others. It is more than likely that the functional variable(s) is different for each individual child. Although anxiety is a common construct invoked in mainstream psychology, behavioral conceptualizations of anxiety-related disorders are not yet well established empirically or theoretically since anxiety as a term encompasses such a broad array of responding (Dymond & Roche, 2009; Friman, Hayes, & Wilson, 1998). What is clear, however, is that over time a child learns alternative responses to speech (i.e., silence) in the presence of others and/or in the presence of particular situations and settings.

Treatment for selective mutism has typically focused on behavioral and pharmacological interventions. This article focuses primarily on behavioral interventions (see Carlson, Mitchell, & Segool, 2008 for a review of pharmacological treatments for selective mutism). Largely due to the low prevalence of selective mutism, research on behavioral interventions has typically employed case studies and single-subject designs. In addition, although anxiety is often invoked as an aspect of the conceptualization of why children are selectively mute, behavioral interventions have largely focused on the operant components. Shaping procedures and stimulus generalization procedures (also termed programming common stimuli and stimulus fading in previous research) have been identified as the two most common effective behavioral interventions for promoting speech (Stone, Kratochwill, Sladezcek, & Serlin, 2002).

Shaping and stimulus fading are two very different treatments. Shaping involves reinforcing successive approximations to a target behavioral goal, often speaking in the presence of others. In short, the target goal for treatment in a stimulus fading procedure focuses on the content. For a child with selective mutism, the target goal for shaping is focused on what the child does to communicate. Stimulus generalization involves reinforcing verbal communication in the presence of different stimuli, such as new peers, adults, or different environments, such as new situations or settings. In short, the target goal for treatment in a stimulus fading procedure that focuses on the context in which the child communicates. For a child with selective mutism, the target goal for stimulus fading treatment is focused on where, when, and/or with whom the child communicates. There are similarities between the two treatment approaches. Both forms of treatment typically involve the systematic prompting of communicative responses (vocal or non-vocal) and the subsequent fading of those prompts.

Functional Behavior Assessment: Interviewing and Observational


Prior to designing and implementing a behavioral intervention for the treatment of selective mutism, many decisions must be made such as determining the verbal behavior to be shaped (e.g., goal, baseline, and successive steps), what prompts to use, how prompts will be faded, what stimuli (e.g., teacher, other student) need to be introduced and how stimuli will be introduced (i.e., procedure for fading), and what reinforcement and reinforcement schedule will be used. These treatment decisions must be individualized to effectively address each child's needs and current functioning. It can readily be argued that effective treatment decisions are most likely to occur when guided by relevant assessment data. Decisions regarding behavioral interventions, such as shaping or stimulus generalization, often rely on data from behavior assessment such as interviews and observations from both clinical and naturalistic settings. Although some behavioral rating scales specific to selective mutism exist (e.g., Hooper & Linz, 1992), there does not appear to be widespread use of rating scales in the treatment literature. Such scales may be more commonly used as part of a diagnostic or screening evaluation.

For treatment purposes, behavior observation has been described as the "key lynchpin" of assessment and treatment of selective mutism (Kearney & Vecchio, 2006, p. 144). Behavior observations are typically guided by information gathered from interviews with significant caregivers in the child's life (i.e., parents, teachers). Kearney and Vecchio (2006) suggest several procedures for the behavioral assessment of selective mutism, including the use of key questions for interviewing parents and structured observations in multiple settings such as the classroom, playground, and home (see Table 1). Additionally, the use of daily logs to collect data from parents, teachers, and/or children on speaking is recommended. The procedures described by Kearney and Vecchio appear very promising for guiding assessment of children with selective mutism. What is missing, however, are specific guidelines or steps for how to analyze and use these data to inform treatment decision-making.

Some years ago, Labbe and Williamson (1984) provided some important guidance for linking assessment and treatment. Specifically, they suggested five links between assessment outcomes and treatment strategies that were based on a review of selective mutism research at that time. Table 2 presents an adaptation of Labbe and Williamson's treatment suggestions. Conceptually the links between assessment data and treatment decisions prescribed by Labbe and Williamson seem to make sense, but there has not been any follow-up research or other literature that has described and applied those guidelines for linking assessment and treatment for children with selective mutism.

This article expands upon the linkages between assessment and treatment previously outlined by Labbe and Williamson (1984) and the assessment procedures described by Kearney and Vecchio (2006) by more specifically describing guidelines for conducting behavior observations of children with selective mutism and then explicitly linking the data from those observations to the design, implementation, and monitoring of behavioral treatment for children with selective mutism. Case studies are presented as examples of the potential treatment utility of the assessment/observation process. Finally/ some suggestions for future research to establish evidence-based assessment protocols to facilitate the effective behavioral treatment of selective mutism are provided.

Functional Behavior Observation of Children with Selective Mutism

For the purposes of treatment planning for selective mutism, behavior observation is conducted for two primary purposes; 1) to identify the settings and situations in which the child does and does not speak, and 2) within each setting and situation, to identify what the child does to communicate. The first purpose refers to where, when, and with whom, the child communicates and the second purpose refers to what the child does to communicate (i.e., whispering, nonverbal gesturing, vocalizing, etc.) within each of the first three contexts. This has been referred to as the AB (Antecedent-Behavior) model of functional analysis (Hanky, Iwata, & McCord, 2003).

Typically, in a functional behavior assessment a third purpose is also identified, namely to identify the relevant consequences for speaking and not speaking. Consideration of the third purpose has been referred to as the ABC model (Hanley, Iwata, McCord, 2003). Identifying the consequences of a behavior, such as silence, is difficult. One could look to identify consequences for the behaviors that take place instead of vocalizations (e.g., non-vocal behavior), but this has not been accomplished to this point. Also, one could seek to identify the consequences that function to maintain vocal behavior in the presence of other stimuli. This has also not been demonstrated in previous research. There has been preliminary work on systematically manipulating conditions as part of a brief experimental analysis (i.e., analog observations using the ABC model of functional analysis) to determine conditions in which a child will and will not speak (Schill, Kratochwill, & Gardner, 1996). Such an approach would have the advantage of improved time efficiency relative to the time involved with conducting naturalistic observations. However, such an approach requires control conditions in which the child will speak or otherwise communicate in order to have comparisons of differences in rate of speech between experimental conditions. Additional research using brief functional analyses with children with selective mutism is needed before such an approach can be recommended for common practice. For this reason the AB model of functional behavior assessment is the focus of this article.

At least the first two objectives must be met in order to effectively design a treatment that focuses on the individual child's use of communication skills in different environmental settings, with different people, and across different situations and times. By completing a functional behavior assessment of a child's baseline functioning, an individualized treatment program with a reasonable treatment hierarchy and stepped treatment goals can be developed.

Where (Settings)

The hallmark characteristic of selective mutism is that a child speaks in some settings, typically the home, but not other settings, such as the classroom or in public places such as restaurants. One of the first steps in assessment would be determining all the settings common to a child's daily routine and determining those settings in which the child speaks and those settings in which the child does not speak or otherwise communicate. This can largely be accomplished through interview with the parent and significant others (i.e., teachers). Subsequent observation of the child should take place in a setting in which the child talks readily and, of course, in settings where the child does not talk. Sometimes video recording of the child in these settings may be used if the clinician is not able to arrange for direct observation. Obtaining a sample of typical speech in a setting where speech typically occurs is essential for assessing the child's verbal skills and to rule out an underlying speech or language disorder or other diagnosis.

When (Activities/Demands)

Within settings there are situations defined by activities in which the child participates, and/or demands or expectations to which the child is expected to respond. There are likely to be particular activities in which the child is more likely to talk compared to other activities. It may be that the child more readily talks during a play activity in the classroom compared to when called upon to answer a question. It may be that a child is more likely to talk while playing soccer during recess compared to playing "Red Rover." In other words, once the settings are determined in which a child talks and does not talk, it is helpful to begin to more specifically analyze activities that occur in those settings in which the child is more or less likely to talk. Assessing the specific variables inherent with different activities/demands is important to understanding how these different variables within a given setting may affect a child's speaking. For example, it may be that certain variables such as tangible objects, preferred activities, or needs, (i.e., full bladder), are more likely to be situations in which the child speaks, whereas other variables such as taking turns reading aloud in class are not likely to produce talk. It may be that intervention will include programming those variables or stimuli (i.e., activities/demands) likely to produce speaking in settings where speaking is less likely to occur.

With Whom

Similar to identifying the different activities and demands within settings in which the child is more or less likely to talk, it is also helpful to identify particular adults, children, and community members with whom the child is more or less likely to talk. Our experience has been that a child may be likely to have one or more peers in the classroom with whom he or she may communicate in some capacity (vocally or non-vocally), and that it is the adults in the classroom (teachers, paraprofessionals) with whom the child is less likely to communicate. It is important to identify all individuals with whom the child is most likely to talk or communicate and those with whom the child is least likely to talk and/or communicate.

What (Vocal and Non-vocal Communication Repertoire)

The ultimate goal of treatment is for the child to be speaking across all settings, situations, and with all relevant individuals. Therefore, it is important to determine initially the child's typical communication skills in settings in which the child typically talks. It is also important to determine across all other settings and situations how, or if, the child typically communicates if the child does not talk. For example, does the child use gestures, pictures, whispers, mouth words, use others as interpreters, or refrain from communication altogether? The assessment of current communication skills, particularly in settings in which talk is limited, will help determine the starting point for intervention. For example, if a child refrains from any communication (i.e., "freezes" when attended to by others) than the first step may be to prompt and reinforce any type of communicative response. If the child occasionally responds non-vocally, but at a very low rate in particular situations or settings, then the first step may be to increase the rate of non-vocal communication prior to attempting to shape approximations to speaking. If the child occasionally whispers in some settings or situations, then the first step may be to increase the decibel level of speaking. In sum, it is important to identify the type and rate of communication within the where, when, with whom analysis so that the clinician knows what communication skill needs to be addressed first in treatment.

Observation Procedures

A copy of a basic observation form is shown in Figure 1 with recommended definitions shown in Table 3 and described below. The settings in which observations would occur are determined based on parent and teacher interviews. Observation duration would typically last the extent of the targeted activity (e.g., 20 to 60 minutes). As with most types of observation, a minimum of two observations in a particular setting and situation to are needed to estimate variability in data. Multiple observations may be especially important in cases of selective mutism as a child's reactivity to an observer may be particularly enhanced. Naturalistic observations are recommended, although clinic rooms with one-way mirrors are also helpful.

Operational Definitions for Observation Coding.

A 15-second partial interval recording system is recommended for coding behavior. Four basic behavioral codes are recorded within two categories; 1) Communication behavior and 2) Stimulus condition. The operational definitions for the behavior codes are provided in Table 3 for use in classroom settings and can be used in conjunction with Figure 1. Although in current form the definitions target a classroom context; the definitions could readily be tailored for any setting or activity observed.

The primary point is to observe and record the type of communication exhibited and the opportunities the child is provided to communicate in the situation being observed. It may be that there are situations when the child is not communicating, but the child is provided few if any opportunities to communicate in those situations. Thus, intervention may seek to alter the environment to increase the frequency of opportunities to communicate or at the very least make available opportunities more salient (i.e., discriminable) to the child (and observer). However, as currently written, "Prompts" (or opportunities to respond) are probably the least well-defined aspect of the behavior code. Determining a "prompt" to communicate requires some clinical judgment on an observer's part that a communicative response is "expected" in a given situation or interaction. Again, if it is not clear what opportunities or prompts are in place for a child to communicate, then one of the key aspects of treatment may be making prompts for communication more explicit both for the child, but also for the person initiating interaction and for the observer. Two case examples are presented to illustrate the use of interviewing and observation in developing effective interventions for selective mutism.

Functional Behavior Assessment Informing Treatment Decisions: Case Examples

The following two children were seen at an outpatient psychology clinic in a Midwest urban city. After interviewing the parents and observing each child in the clinic and at school, each child was given a diagnosis of Selective Mutism.

Case 1: Darren

Darren was a 10-year-old Caucasian male who attended a regular education public school fourth grade classroom in a rural school district. He lived with his biological mother and her parents. He had regular visitations with his biological father. There was a reported history of anxiety on both parents' sides of their extended families. Darren had a long history of not talking with teachers or peers at school or with adults, including family members, and peers in settings outside the home. He spoke with his mother, his father, and his grandmother, but not with his grandfather at home. He had been prescribed 10 mg. of Lexapro prior to being seen in the outpatient clinic and there were no changes in medication throughout services. Darren was described as compliant at home and school. No concerns were noted with academic skills.

Two baseline observations were conducted at the school using 15-second partial interval recording and the behavior codes described earlier. Observations occurred during small group reading, recess, lunch, independent reading, and science which included large group instruction and small group projects. Observations were conducted across settings in the school environment to determine if there were specific activities and demands and people within those settings that appeared to affect the probability of communication.

The results of these two baseline observations are shown in Table 4. In the top half of the table (first two rows) are the percentage of intervals during the observation in which Darren was provided opportunities from his teacher or from peers to respond vocally or non-vocally. In the bottom half of the table (bottom five rows) are the percentage of intervals that Darren did respond when provided the opportunity. During the first observation that included small group reading consisting of Darren and three other peers (duration of 23 minutes), the teacher asked the group questions during 50% of intervals and directly asked Darren questions during 16% of intervals observed (see first column of table). He responded non-vocally (i.e.,. nodding his head yes or no) to 11% of the opportunities provided directly to him. In short, out of the 16% of the intervals in which he was provided an opportunity to respond, he responded non-vocally in 11% of the intervals. He did not speak throughout the observation. His teacher reported the observation was typical of his behavior. She had taught him for 3 years and he had never spoken in the class, although he reportedly responded non-vocally at times when directly prompted.

The second set of observations had similar findings. Darren did not respond vocally at recess (17.5 minutes), lunch (20 minutes), independent reading (21 minutes), or during science (31 minutes). He provided minimal non-vocal responses to some of the opportunities provided by his peers and teacher. The observation data indicated that when Darren was provided opportunities from the teacher or peers to respond, he would respond non-vocally approximately 33% to 37% of the time. It was clear that there were few opportunities to speak (see top two rows of Table 4). It is possible that the teachers, staff, and other students no longer had much expectation for Darren to speak and did not provide very many opportunities to speak.

Based on these data, it was decided that Darren would benefit from intervention that initially targeted providing more opportunities to speak and increasing the frequency of non-vocal responses. This treatment plan was developed based on analysis of observational data demonstrating minimal prompted communication (low rates of non-vocal responses and no speaking) across most environments and people. These data led to consideration of Labbe and Williamson's recommendation that "If child speaks to no one in all environments then some type of response initiation is needed ..." (Table 2). To do this, it was determined that it was first important to increase the number of opportunities Darren was provided to speak in the classroom. Given the low rates of any communication that were occurring, it was decided first to increase the rate of non-vocal communication that occurred in response to increased opportunities. Once non-vocal communication was occurring more frequently, then intervention could focus on shaping the form of communication from non-vocal to vocal. Note here how the functional behavior assessment helped facilitate the decision-making process relative to establishing the conditions necessary to implement effective treatment If shaping of speaking had been attempted first, it may well have been unsuccessful as there were so few opportunities for Darren to speak.

In consultation with the school team and caregivers, it was decided that the teacher would provide direct questions to Darren during small group reading time. Although assessment data indicated that there may have been more opportunities in science for Darren to respond, science was a whole class activity. The small group reading activity allowed for more control by the teacher of the questions to be asked and the smaller group provided more opportunities for individual attention to be provided to Darren. Darren was given a criterion of responding non-vocally (e.g., nodding or pointing) three times during small group reading to earn a reward at home. The reward consisted of a small grab bag of items/privileges chosen by Darren and his mother. Darren could pull one reward from the grab bag only on the days that he met criterion for reward. His teacher communicated to his mother using a school-home note. For a week prior to initiation of the intervention, Darren practiced with his mother at home and with the speech/language therapist at school, responding non-vocally to questions while reading. As Darren demonstrated success with the intervention, the criterion for earning his reward was increased (i.e., five and then seven non-vocal responses). Darren's non-vocal responses to teacher prompts during small group reading increased from a baseline of 11% to 78% over the course of approximately five weeks.

Although the intervention was targeted to small group reading, opportunities to respond to the teacher were observed in subsequent observations to increase across class activities (targeted environmental change) and Darren's non-vocal responses to teacher questions were observed to increase in frequency across classroom activities (targeted behavioral change). For example, Science was targeted next and within a few weeks, Darren was non-vocally responding in science and reading to 81% of opportunities provided. By the end of the school year (approximately 5 months) Darren was responding non-vocally (including some mouthing of words with no sound) to almost all opportunities from the teacher and peers, suggesting generalization of this skill.

He continued not to exhibit speaking. In the summer, his family moved and he transferred to a new school for the new school year. Observation in the new classroom indicated that non-vocal responding was still high and frequent opportunities to respond were present with peers and teacher. Consistent with earlier consideration of Lab-be and Williamson's guidelines, shaping was considered as the most promising intervention. The intervention was changed to reward only speaking in response to teacher questions. Shaping of speaking was included in the intervention as whispering was initially allowed. Once he was successfully speaking in whispers, the expectation for reward was increased to a more typical voice decibel. After he was speaking in response to teacher questions, the intervention was changed to reward initiation of questions by Darren to the teacher. Darren was speaking to the teacher 80% of opportunities provided. He was readily talking with peers. When criterion to earn reward was changed to asking questions of the teacher, he improved to asking at least one question daily. This process took about three months.

Observation and teacher report indicated that Darren continued to progress with increasing frequency of speaking to peers and teachers with the reward program. Note that the primary intervention throughout was shaping frequency of communication responses and form of communication response. He eventually was communicating freely with students and the reward system was discontinued.

In the next case study, the initial functional behavior assessment suggested that a different type of intervention was appropriate.

Case 2: Alex

Alex was a seven-year-old Caucasian male in second grade in a suburban public elementary school. He lived with both biological parents and two siblings. He reportedly talked freely at home. He rarely talked in public settings with adults, including his parents, and had rarely spoken with adults at school. There were no reported health concerns and he was not prescribed, any medication. No concerns were noted regarding academic skills. No concerns were noted for compliance at home or at school. Alex was reported to talk with peers in the classroom, but not so the teacher could readily hear him. His teacher reported that he would whisper to peers in her presence. He would respond non-vocally to his teacher (e.g., nodding, gestures). The school counselor noted that he had responded to her with one to two spoken word answers on occasion during a one-on-one conversation. In addition, Alex was reported to occasionally respond to the physical education teacher with one word spoken answers.

In early fall of the school year, two observations were conducted in the classroom using 15-second partial interval recording and the observation codes mentioned above. Activities observed at school included center time (8.5 minutes), spelling (23 minutes), reading (15 minutes) and lunch (14 minutes). Physical education was led by a different teacher. A summary of findings from the first observation is presented in Table 5. Alex was observed to communicate readily with peers both with and without speaking. He responded to opportunities from the teacher non-vocally, but was not observed to speak to her. Data from a second observation compiled across science (15 minutes), math (25 minutes), and physical education (10 minutes) are also shown in Table 5. The data from the second observation are similar to the first in that Alex communicated readily with peers both vocally and non-vocally. He responded non-vocally to the teacher. He spoke to his physical education teacher on one occasion. That teacher reported that Alex readily talked with peers in his presence during physical education, excelled at athletics and appeared comfortable in physical education.

A meeting was held with Alex's school team including his parents, teacher, school psychologist, school counselor principal and consultant. Based on the information and data collected, it was determined that Alex was speaking to peers in the school setting, but not consistently to adults. These data led to consideration of Labbe and Williamson's recommendation (Table 2) that "If child speaks to only one or a few people in most environments then consider stimulus fading of new people in target environments." Alex spoke to peers across most activities within the school setting, but did not speak with his teacher or other adults. Therefore, it was decided to implement a stimulus fading intervention with his teacher as the changing stimulus. As the functional behavior assessment data indicated, physical education was certainly a setting/activity that could be utilized to introduce the teacher's presence. 'But, this was not a time the teacher was readily available and it was difficult to manipulate steps in the stimulus fading process with the physical education activities. Also, there were fewer opportunities for Alex to respond during this activity. Since lunch also appeared to be a time when Alex talked readily with peers, and there were plenty of opportunities already embedded in this activity for Alex to respond, it was planned that his teacher would slowly introduce her presence into this situation. Steps in this stimulus fading procedure were carefully planned to ensure gradual exposure of the teacher.

Unfortunately, his teacher found she had less time than expected over the lunch period to be available with the children and only attempted to implement the intervention weekly. She also may have moved or inserted herself too quickly into the group with whom Alex was sitting and Alex did not speak in her presence. Upon further consultation with the school team, it was subsequently decided to conduct stimulus fading with the school psychologist. While the ultimate goal was speech with the teacher, expanding Alex's use of speech in the school setting with any adult was considered an important treatment compromise and it was expected that introducing the school psychologist into an activity where Alex was speaking to peers would facilitate other stimulus fading procedures. Therefore, the focus of the stimulus fading intervention changed to having the school psychologist work with Alex and peers in a small group and the school psychologist gradually increased her presence as part of this small group. This activity took place in the school hallway. Following implementation of this program, Alex's school psychologist reported that he talked spontaneously with her during the small group intervention in the hallway.

In addition, a second stimulus-fading program was started with the classroom teacher at a time when she had fewer demands and time constraints. Alex's mother began bringing him to school early to read and play games in the classroom in the presence of the teacher. The teacher was asked to increase her proximity to Alex and his mother over time. Alex's mother reported that he talked and read aloud with her while his teacher was in close proximity. Once he began talking more frequently to his mother in the presence of his teacher, a reward system was implemented to increase the frequency of speaking to his teacher consistent with Labbe and Williamson's recommendation that "if the child speaks occasionally to most people in most environments then consider contingency management ..." (Table 2). The reward system consisted of a grab bag procedure similar to the one described for Darren. Follow-up observation in the classroom indicated Alex's oral responses to the teacher increased to 54% of opportunities provided by the teacher. This was considered a substantial improvement as he was not vocally responding to the teacher prior to intervention. Additionally, Alex performed parts in "Readers Theatre" (a small group play) for several teachers and students at his school. Finally, toward the end of the school year, further stimulus fading was conducted by gradually introducing Alex's teacher for the next year into the small group led by the school psychologist and Alex began talking spontaneously with this teacher.


It is important to note that these are therapeutic case studies with only a few repeated observational measurements, and the data are meant primarily to support particular points respective to linking functional behavior assessment to intervention. The data are not meant, and should not be used, as empirical demonstrations of the efficacy of these particular interventions. However, the data are consistent with what may typically be collected as part of the day-to-day activities by educators and school psychologists to measure the effects of an intervention. Also, this article focuses almost solely on the observational process in keeping with Kearney and Vecchio (2006) that observation is the "key linchpin" to effective treatment for children with selective mutism. Interview data are highly important in guiding observation and providing additional information to guide treatment decisions, but interview data alone are not typically reliable and valid for treatment decisions (McConaughy, 2000). Using the observational and interview data collected following a treatment referral, it was possible to use data-based decision making to guide intervention design in each of these cases.

Linking Assessment Data to Treatment

As illustrated by the cases presented, consideration of the 4 Ws: Where (settings), When (activities and demands), With Whom (people), and What (vocal and non-vocal communication), as a structured part of an interview and observation process provides the data to make decisions regarding effective treatment for children with selective mutism. The resulting data can then be meaningfully linked with effective intervention as recommended by Labbe and Williamson (1984) and Stone et al. (2002). Psychologists working in schools with children with selective mutism are advised to complete a functional behavior assessment by gathering observational and interview data across settings, activities, and individuals with consideration for the types of communication the child exhibits and the opportunities provided to the child to communicate.

Consider Opportunities Available

As noted in the case examples presented here, it is very important to consider the opportunities present for children to respond. Assessing the opportunities to vocally or non-vocally respond has not been previously discussed in any of the literature on assessment or intervention for children with Selective Mutism. If a child does not have opportunities to communicate, there is no reason to expect that the child will communicate. Without opportunities or prompts to communicate, communication largely becomes a response by the child due to motivating operations that may be more difficult to assess (e.g., hunger, full bladder). Assessing the frequency, or rate, and/or type of opportunities available within the contexts, settings and activities the child vocally and non-vocally communicates is vitally important to designing effective treatment and monitoring progress.

AB vs. ABC Functional Behavior Assessment

The emphasis of this article as been on the assessment of antecedent-behavior relations. Evaluating the reinforcing and punishing consequences of Selective Mutism may be very difficult as this is a behavior that is defined by its antecedent relations (i.e., occurs selectively in the presence of some stimuli but not others). The consequences maintaining these stimulus-behavior relations may be negatively reinforcing (i.e., escape from aversive stimuli) or positively reinforcing (i.e., increased social attention or access to preferred items and activities). On the other hand, vocal behavior may be punished in the presence of some stimuli (i.e., social attention, expectation to complete a task). Identification and manipulation of punishing or reinforcing stimuli for mute or vocal behavior is difficult and an area that requires additional consideration and research. It is possible that assessment of the ABC contingency will lead to more effective treatment than current AB assessments.

Future Research Directions

Additional empirical support is needed to demonstrate the treatment validity of the functional behavior assessment procedures described here for designing treatments. Such treatment validity evidence may come from carefully controlled single-subject experimental designs demonstrating the effectiveness of treatment following detailed assessment procedures including observation and interviews or from other experimental designs tailored to clarify the effects of different treatment decisions based on individual clients' assessment data using this type of observation process (Hayes, Nelson, & Jarrett, 1987).

Also, other types of observational procedures that assist with decision-making regarding treatment need to be developed and evaluated for their efficacy and efficiency. The current recommended observation procedure uses interval recording. We chose partial-interval recording primarily for ease of use by the observer and to avoid trying to discern discrete events of communicative behavior as part of a frequency count. Partial interval recording may overestimate actual responding (Johnston & Pennypacker, 1993). Given the low rate of responding inherent in selective mutism, this was not perceived as a problem, however, the clinician/educator needs to be aware of the advantages and disadvantages of any observational procedure being used.

In classroom settings, it is not always clear what should be the expectations for frequency or rate of speaking. Gathering data on classmates' speaking as comparisons may be beneficial in determining the discrepancy between a child's current and expected communication behavior. In addition, it may be helpful to have information on the frequency or rate of opportunities typically provided to other children in the classroom to speak. Additional research on the most effective and efficient methods of observation of communicative behaviors is needed.

Researchers generally are most interested in determining the efficacy of various treatments, and continued research on the efficacy of treatments for selective mutism is clearly needed. However, it is also imperative that research on valid assessment processes linked to effective treatment decisions be conducted to help guide practitioners in effective data-based decision making. It is hoped that this article in attempting to delineate more clearly the link between observation and treatment decisions will help facilitate additional research on this topic.


American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed, text revision). Washington, D.C.: Author.

Bergman, R.L., Piacentini, J., & McKracken, J.T. (2002). Prevalence and description of selective mutism in a school-based sample. Journal of the American Academy of Child and Adolescent Psychiatry, 47, 938-946.

Black, B., & Uhde, T. (1995). Psychiatric characteristics of children with selective mutism: A pilot study. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 847-856.

Carlson, J.S., Mitchell, A.D., & Segool, N. (2008). The current state of empirical support for the pharmacological treatment of selective mutism. School Psychology Quarterly, 23, 354-372.

Dymond, S., & Roche, B. (2009). A contemporary behavior analysis of anxiety and avoidance. The Behavior Analyst, 32, 7-27.

Elizur, Y., Perednik, R. (2003). Prevalence and description of selective mutism in immigrant and native families. A controlled study. Journal of the American Academy of Child and Adolescent Psychiatry, 42, 1451-1459.

Friman, P.C., Hayes, S.C., & Wilson, K.G. (1998). Why behavior analysts should study emotion: The example of anxiety. Journal of Applied Behavior Analysis, 31,137-156.

Hanley, G.P, Iwata, B. A., & McCord, B.E. (2003). Functional analysis of problem behavior: A review. Journal of Applied Behavior Analysis, 36,147-185.

Hayes, S.C, Nelson, R.O., & Jarrett, R.B. (1987). The treatment utility of assessment: A functional approach to evaluating assessment quality. American Psychologist, 42, 963-974.

Hooper, S.R., & Linz, TD. (1992). Elective mutism. In S.R. Hooper, G.W. Hynd, & R.E. Mattison (Eds.), Child psychopathology: Diagnostic criteria and clinical assessment (pp. 409-459). Hillsdale, NJ: Lawrence Erlbaum.

Johnston, J.M., & Pennypacker, H.S. (1993). Strategies and tactics of behavioral research, 2nd Ed. Hillsdale, NJ: Lawrence Erlbaum.

Kumpulainen, K., Rasanen, E., Raaska, H., & Somppi, V. (1998). Selective mutism among second graders in elementary school. European Child and Adolescent Psychiatry, 7, 24-29.

Kearney, C.A., & Vecchio, J. (2006). Functional analysis and treatment of selective mutism in children. Journal of Speech-Language Pathology and Applied Behavior Analysis, 1(2), 141-148.

Labbe, E.E., & Williamson, D.A. (1984). Behavioral treatment of elective mutism: A review of the literature. Clinical Psychology Review, 4, 273-292.

McConaughy, S.H. (2000). Self-reports: Theory and practice in interviewing children. In E.S. Shapiro & T.R. Kratochwill (Eds.) Behavioral Assessment in Schools: Theory, Research, and Clinical Foundations, 2nA Edition. New York, NY: Guilford Press.

Schill, M.T., Kratochwill, T.R., & Gardner, W.I. (1996). An assessment protocol for selective mutism: Analogue assessment using parents as facilitators. Journal of School Psychology, 34,1-21.

Stone, B.R, Kratochwill, T.R., Sladezcek, L, & Serlin, R.C. (2002). Treatment of selective mutism: A best-evidence synthesis. School Psychology Quarterly, 17,168-190.

Vecchio, J.L., & Kearney, C.A. (2005). Selective mutism in children: Comparison to youths with and without anxiety disorders. Journal of Psychopathology and Behavioral Assessment, 27, 31-37.

Mark D.Shriver Munroe-Meyer Institute for Genetics and Rehabilitation Natasha Segool University of Hartford Valerie Gortmaker Applewood Centers, Inc., Cleveland, Ohio

Correspondence to Mark D. Shriver, Ph.D, Associate Professor, Psychology Pediatrics Munroe-Meyer Institute, University of Nebraska Medical Center 985450 Nebraska Medical Center Omaha, NE 68198-5450; e-mail:
Table 1 Interview and observation techniques for use in
designing treatment

Interview Questions                    Observation Techniques


What settings occasion a         Narrative recording of setting
child's mutism                   variables
(e.g., home, school, community   associated with speech and mutism

How long has the mutism
occurred in each setting?


With whom will the child speak   Record who the child communicates
Freely                           with spontaneously or following
or become mute?                  a prompt. Record prompts that go

When mutism occurs in each       Record who is in the immediate
situation,                       environment when speech or mute
is the child alone or with       behaviors occur.

                  Quality of Communication

How is the child's mutism        Record the number of words spoken
manifested                       or
in each setting?                 type of nonverbal communication

What compensatory behaviors      Record compensatory behaviors
the child show to communicate    (e.g., whispering, pointing,
with                             nodding,
others?                          mouthing, crying, frowning,
                                 temper tantrum, pushing, or
                                 Narrative recording of child's
                                 and communication skills in
                                 relation to
                                 same-aged peers

                  Antecedents and Consequences of Mutism

What are the specific            Record key antecedents (e.g./
antecedents and                  demands/
circumstances that surround      expectations of situation or
each                             social
instance of mutism/speech?       approaches from others)

How do others respond to a       Record key consequences (e.g.,
child's                          parent or teacher acquiescence;
mutism (e.g., ordering food or   accommodation of a child's mutism)
completing tasks for the child;
whispers in the ear or
rearranging a setting to
accommodate a
child's mutism)?

Can the child be enticed to      Record any possible reinforcer or
speak                            change in environment used to
audibly in these situations in   produce
any way?                         speech.

To what does the child or the    Record evidence of anxiety as
family                           indicated
or the teacher attribute mutism  by visible arousal symptoms (e.g.,
oppositional, anxiety, skill     blushing, body tense, eyes cast
deficit)?                        downward), escape, withdrawal,
                                 or avoidance behaviors

Adapted from Kearney & Vecchio, 2006, p. 143-4

Table 2 Treatment recommendations linked to assessment outcomes

Recommendation    Speech     Person     Setting       Treatment Steps
                 Frequency  Variable   Variable

1               Limited/    Most      Most          A) Contingency
                occasional  people    environments  Management for all

2               Typical     One or    Most          A) Stimulus Fading
                rate        limited   environments  of new people in

                Typical     people    One           A) Stimulus Fading
3               rate        Most      environment   of environmental
                            people                  stimuli

4               Limited/    One or    One           A) Stimulus Fading
                occasional  limited   environment   of new people into
                            people                  environment where
                                                    speech already
                                                    occurs and
                                                    B) Stimulus fading
                                                    of environmental

5               No speech   No        No            A) Response
                            people    environments  initiation, i.e.
                                                    modeling, or
                                                    response cost and
                                                    B) Stimulus fading
                                                    of new people

Note: In all treatments described above, contingency management
procedures focused on positive reinforcement for speech,
reinforcement fading, and maintenance procedures were also
proposed following the implementation of the primary treatment.
Adapted from Labbe & Williamson, 1984, p. 289

Table 3 Operational Definitions of Observational Codes

Variable                    Definition

                Communication Behavior

Non-vocal       Nodding or shaking head or gestural
(NY)            response with hands, arms, or shoulders
                (e.g., motioning/pointing with
                hands, shrugging shoulders) to initiate
                communication or respond to prompt to

Vocalization    Using voice (any volume) to
(V)             initiate communication or to
                respond to prompt to communicate.

                Stimulus Condition

Opportunity     Teacher or peer communication (NV or
(prompts)       V as defined above) with target student
                (or group in which target student is included)
                prompting target student to
                communicate with either vocalization
                or non-vocal behavior.

Initiation (no  Target student speaks or non-vocally
prompts)        communicates with peer or
                teacher (or other) without observable prompting.

Table 4
Percent of intervals Darren was provided opportunities to
communicate and percent of opportunities he did communicate
(all percentages are rounded).

Variables          Small Group  Recess  Lunch  Independent  Science
                     Reading                     Reading

Opportunities                0       4      0            0        2
from peers

Opportunities               16       0      0            0       24
from teacher/

Non-vocal                  N/A      33    N/A          N/A       33
response to peers

Non-vocal                   11     N/A    N/A          N/A       37
response to

Vocal response             N/A       0    N/A          N/A        0
to peers

Vocal response to            0     N/A    N/A            0        0

Non-vocal                    0       0      3            0        o

Vocal Initiation             0       0      0            0        0

Table 5 Percent of intervals Alex was provided opportunities to
communicate and percent of opportunities he did communicate
(all percentages are rounded)

Variables         Centers  Spelling  Reading    Lunch    Observation 2

Opportunities          21         0        0        100              2
from peers

Opportunities           0         8        0        N/A             16
from teacher/

Non-vocal              86       N/A      N/A        100            100
response to

Non-vocal             N/A       100      N/A        N/A             30
response to

Vocal response          0       N/A      N/A        100             80
to peers

Vocal response        N/A         1      N/A        N/A              0
to teacher/staff

Non-vocal               0         0        0  100 (with             45
initiation                                       peers)

Vocal Initiation        0   5 (with        0  100 (with   2 (with peer
                             peers)              peers)         and PE

Note: Observation 2 data were compiled across science, math,
and physical education
Gale Copyright:
Copyright 2011 Gale, Cengage Learning. All rights reserved.