Printer Friendly

Promoting speech in selective mutism: experimental analysis, differential reinforcement, and stimulus fading.

Selective mutism (SM) is characterized as a consistent fail-sure to speak in specific social situations in which the use of speech is expected (Krysanski, 2003). This failure to speak is only characterized as SM when speech is used to communicate in certain situations and not in others (Manassis, Fung, Tannock, Sloman, Fiksenbaum, & McInnes, 2003). According to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; American Psychiatric Association, 2000), SM is classified under the category of disorders usually first diagnosed in infancy, childhood, or adolescence with the qualifying statement that the SM is not due to a lack of knowledge or comfort with the spoken language that is required for the situation, is not accounted for by another type of communication disorder, and does not occur exclusively with pervasive developmental disorder (PDD), schizophrenia, or psychotic disorder (American Psychiatric Association, 2000). An additional diagnostic characteristic of SM is that it interferes with educational or occupational achievement and that the interference is present for at least one month (excluding the first month of school). Thus, if early intervention is not successful, there is a strong possibility that SM can negatively affect success at school, as audible verbalizations are often required to assess current levels of knowledge as well as progress on newly acquired skills.

Several behavior therapy strategies have been used successfully in the treatment of SM including reinforcement contingencies, stimulus fading, modeling, token economies, shaping, prompting hierarchies, systematic desensitization, and mild adversives (Giddan, Ross, Sechler, & Becker, 1997; Nolan & Pence, 1970; Schill, Kratochwill, & Gardner, 1996; Watson, & Kramer, 1992; Rye & Ullman, 1999). In the majority of these studies, no one single behavioral treatment was successful in isolation and the researchers either created a treatment package, utilized multiple phase changes building on previous phases, used an alternating treatment design, or some combination of the above to evoke speech.

In addition to the behavioral strategies described above, differential reinforcement has been used effectively for individuals with SM to shape a behavioral progressions toward vocalizations For example, Giddan, Ross, Sechler and Becker (1997) used this procedure to shape the following behavioral progressions: (a) written messages (b) gestures, head nods and pantomime (c) private tape recordings (d) soft whisper (e) loud whisper (f) vocalizations of animal sounds, coughing, and kazoo (g) soft voice, and (h) full voice in school and beyond. The authors found that the gradual increases in demands increased the likelihood for success in the treatment of SM.

Although SM is topographically different from the externalizing behaviors observed more frequently in schools (e.g. talking out, out of seat, and aggression), SM behavior (i.e., not talking in specific situations) resembles externalizing behaviors in that it serves a function for the individual. The recognized functions of behavior are attention, escape/avoidance, and access to tangible and sensory reinforcers (Cooper, Heron, & Heward, 2007). However, before the function of a behavior can be determined, its dimensions must be operationally defined. Once defined, the behavior can be assessed reliably and baseline performance rates can be identified. Then, information gathered from teacher and parent interviews, rating scales, and direct observation can be used to formulate hypotheses about its function.

The function of SM behavior can be assessed by using a functional analysis (FA). According to Schill, Kratochwill, and Gardner (1996), the process of conducting a FA can be conceptualized by focus, setting, techniques, and person conducting the assessment. FAs are typically used to link assessment data with an intervention. In previous research involving SM participants, treatments have been described with "sufficient detail and clarity" (Cooper et. al, 2007, p. 17) to support easy replication. However, they have been limited in that the functions of the SM behavior have not been assessed systematically through FA (Schill, Kratochwill, & Gardner, 1996). Consistent with the treatment of more typical overt classroom behavior problems, it is necessary to begin the treatment of SM with a FA prior to implementing the intervention.

Schill, Kratochwill, and Gardner (1996) described the treatment of an 8-year-old girl with SM. They collected data from school and home settings, formed hypotheses regarding the function of the SM behavior, and used an analog assessment to determine the behavior's maintaining variables. The FA was conducted by structuring eight easy and hard demand conditions, including (a) baseline, (b) easy demand (a non-vocal response resulted in access to reinforcement), (c) hard demand (only a vocal response resulted in access to reinforcement), (d) easy demand involving prompt, ignore, and praise for "why" questions, (e) easy reading (only the parent present), (f) hard reading (parent plus unfamiliar observer present), (g) easy contingent play session, and (h) hard non-contingent play session. Results of the FA indicated that the presence of an unfamiliar adult was associated with less speech. Therefore, stimulus fading was selected as an intervention. In the end, this study demonstrated that when the hypotheses formed from indirect and direct data are assessed by FA, intervention can be more effectively tied to the treatment.

The present study adds to the current literature by replicating the effective use of a behavioral approach for promoting speech by children with SM. It also extends the work of previous studies (Schill et al., 1996; Giddan, et al.; 1997; Watson et al., 1992) by illustrating how a brief experimental analysis (BEA) can be used to link assessment data to an intervention strategy more effectively than other forms of assessment alone, such as interviews (an indirect form) or even classroom observations (a direct form).



The participant (known here by the pseudonym "Rose") was a 6-year-old Caucasian female attending a public kindergarten in a rural area of the southeastern United States. Rose's father reported that his daughter's history was negative for birth and medical conditions that could place her at risk for communication disorders. All early developmental milestones, including expressive and receptive language, were achieved on time.

Rose's parents were divorced, and she lived with her father. Her mother reportedly presented with a history of drug addiction for which she was receiving inpatient treatment in a state hospital. Rose had not seen her mother in approximately three years.

At the time of this study, Rose communicated normally at home, but only with three people. She spoke face-to-face with him, with a neighbor's eight-year-old daughter, and with a family friend. Occasionally, she also spoke with these individuals on the phone. Recent videotaped interactions between Rose and her father at home were provided to her pediatrician, her current teacher, and the school psychologist. The conversational language sample included on these tapes and the interview information provided by Rose's father suggested that Rose used age-appropriate speech, language, and communication at home.

At school, Rose was described by her teacher as a "bright child" who did not speak to teachers, school personnel, or peers in any school situation but whose written work suggested she had the potential to excel academically in the classroom. The teacher reported that Rose made eye contact but did not respond verbally to the social initiatives of others. The teacher's concern was that this behavior would compromise Rose's academic success, since oral reading was a prerequisite for passing to the first grade. According to her father, Rose revealed that she did not speak at school because she was 'scared someone would make fun of me."

In consideration of this background information, a team decision was made to initiate behavioral assessment and treatment with a provisional hypothesis that SM was an escape/avoidance maintained behavior (Gresham et al., 2001)


Two procedures were included in this study: (1) a brief experimental assessment (BEA) and (2) an intervention based on the BEA results. The BEA took place in a room within the school's library. The room contained books, two tables, and toys (e.g., stuffed animals, puppets) that Rose could access following successful task completion. The environment was free of distractions. No other children were in the present and there was no hallway noise.

Intervention was provided biweekly at 10:00 a.m. in the same library room as was used for the BEA. During intervention, the room was stocked with items (e.g., books, puppets, play materials, candy) that had been identified as potential reinforcers through a rank-ordered preference assessment based on paternal report. A complete description of the preference assessment is provided below. As during the BEA, the library environment was distraction free during intervention.


Information about confidentiality, the nature of a therapeutic relationship, goals of treatment, and the treatment procedures were discussed with Rose's father prior to the intervention. He then gave his consent. In addition, approval from Institutional Review Board (IRB) at the author's university was secured, and informed consent was obtained from Rose's father to include the data from this project in a research report.


The primary dependent variable was the frequency of audible verbal speech behavior (VSB) produced during the sessions. This behavior was selected in consideration of feedback from Rose's father and her teacher, who considered the use of speech in school important for Rose's academic success and social acceptance. See Initial Assessment, below, for details. Two secondary dependent variables were established, based on the work on Giddan et al. (1997). These behaviors included non-vocal behaviors (NVB) and pre-vocal behaviors (PVB) that served a communicative function.

These target behaviors were defined as follows: 1) NVB included gestural communication such as nodding, head shaking, shoulders shrugging, hand signaling); 2) PVB included sub-vocal non-speech communication such as sighs, animal sounds, popping, and blowing; and 3) VSB included any audible words found in the concise Oxford English Dictionary (2004) and any common forms of slang and colloquialisms. In addition, mimes, audible non-words, and audible utterances were scored (Schill et al, 1996).

Each session was recorded by video and viewed by two independent observers. The observers counted all communication signals (NVB, PVB, VSB), and sessions were summarized by calculating the rate of signals in each modality per minute (i.e., the number of signals divided by the number of minutes in a session). The data were graphed and visually inspected for changes in level, trend, and variability with regard to the three dependent variables (Kazdin, 1982).


Interobserver agreement (IOA) was used to assess the reliability of data collection and coding. Two independent observers coded the signals produced by Rose. One observer collected the data while interacting with Rose, and a second observe collected data while observing the same session on videotape. Agreements between the two observers were divided by the total (agreements plus disagreements) and then multiplying by 100 (Kazdin, 1982). The average IOA for this study was 91% (range of 88-100%) across all sessions. IOA for each communicative category is as follows: NVB was 100%, PVB was 88% (83-100%), and VSB was 98% (97-100%).


Prior to beginning treatment, Rose's father completed an Eyberg Child Behavior Inventory (Eyberg, 1999), an Achenbach Child Behavior Checklist for parents (CBCL) (Achenbach, 1991), Social Skills Rating System (SSRS) (Gresham & Elliot, 1990), and a pre-treatment questionnaire that included a compliance component plus the DSM-IV TR (APA, 2000) criteria for selective mutism. The DSM-IV-TR criteria were added to ensure that Rose was correctly diagnosed while the other scale ruled out other potential causes for the SM behaviors. Rose's teacher completed the Functional Assessment Informant Record for Teachers (FAIR-T) (Edwards, 2002), the SSRS, the CBCL for teachers, and the Sutter-Eyberg Student Behavior Inventory-Revised (SESBI-R) (Eyberg & Pincus, 1999). Each of these instruments were selected to further rule out any behavioral or non-compliance issues and to confirm the diagnosis of SM. All assessments were interpreted by the primary researcher, who was a licensed k-12 school psychologist, and an outside school psychologist mentoring the research. In addition, direct observational data were collected in the classroom by the primary researcher to validate the teacher's observation that Rose was neither vocalizing her needs nor communicating non-vocally.

Following this initial data collection, a preference assessment interview, consisting of rank ordering items, was conducted with Rose's father as the informant. This was done to ensure that the reinforcements used during intervention phase included items that were preferred by Rose.

Rose's father was also asked to assess the acceptability of the proposed intervention prior to the onset of treatment. A modified Treatment Evaluation Inventory-Short Form (TEI-SF) (Lyst, Gabriel, O'Shaughnessy, Meyers, & Meyers, 2005) was used for this purpose. Treatment acceptability is defined as the extent to which treatment is considered reasonable, justified, and fair to consumers, teachers, and caretakers (Lyst et al., 2005). Following treatment, the TEI-SF was given to the parent again to assess posttest acceptability. The results for each of these assessment tools, the preference assessment, and the TEI-SF are presented in the results section of this paper.


Initial data collection was conducted by the author in the classroom to validate the referral question through direct observation of Rose's behavior. Three 30-minute event-recording observations were utilized to document the rate of NVB, PVB, and VSB per minute in the classroom during class instruction across three different days. The rate was determined by the number of each mode of communication divided by the time of the session in minutes.


Based on indirect measures (i.e., interviews, rating scales) and direct measures (classroom observations) of Rose's behavior, a hypothesis was formed regarding the function of Rose's SM behaviors. Specifically, it was hypothesized that Rose's SM behavior functioned as a means of avoiding or escaping social attention from the teacher, peers, and those less familiar with Rose (Gresham, Watson, & Skinner, 2001). To investigate this hypothesis, a brief experimental analysis (BEA) was implement-ed as described below.


The BEA used in this study was adapted from the protocol used by Schill et al. (1996). Its purpose was to assess the hypothesized function of Rose's SM behavior and to determine the appropriate starting point for intervention. Throughout the BEA, questions were posed that required a verbal response from Rose. The questions were categorized as easy and hard. Easy questions were those that asked about content that Rose had been taught within the Kindergarten curriculum and that her father had observed in her repertoire (i.e., identification of name, age, color hair, color eyes, city). Hard questions were those that asked about content within the kindergarten curriculum that had not yet been observed in her repertoire (e.g., site word identification, counting past 20, counting by twos). Note that these definitions of "easy" and "hard" were based on levels of academic difficulty and not on communication modality as in Schill et al.'s (1996) study. This modification was made in consideration of the referral question from the teacher, who had indicated that Rose's SM behavior was compromising her potential for academic success by preventing the assessment of skills, such as reading aloud, which require audible verbal responses.

Two experimental conditions were utilized during the BEA sessions. This resulted in four unique sessions during which to observe Rose's communicative behavior. Each session provided Rose with the opportunity to respond 10 times as 10 questions were posed during each session. The four sessions were distinguished as follows: (1) 10 easy questions with a familiar social partner (father), (2) 10 more questions beginning with easy and moving to difficult questions with a familiar social partner (father), (3) 10 easy questions with a novel social partner (researcher), and (4) 10 easy and difficult questions with a novel social partner (researcher). The dependent variables across all sessions of the BEA were NVB, PVB, and VSB. Note that this BEA differed from a complete FA (Iwata, 1994) in that the consequences were not manipulated based on condition. Instead, this BEA was adapted from methodology discussed by Schill et al. (1996) which manipulated the antecedents as opposed to the consequences.

During the first and second sessions, Rose and her father sat at a table in the room. Also in the room (but out of Rose's line of vision) were the author and a video camera. Rose was informed that the session was being videotaped.

To begin session one, Rose's father presented her with a series of 10 easy questions. The questions were provided to him in advance, together with direct instructions on when to proceed from one to the next. For the purposes of the BEA, two minutes was the maximum time allowed to wait for a response. Questions were not repeated. Figure 1 shows that Rose failed to respond by VSB or PVB and used only a limited amount of NVB. However, she did indicate to her father that she wanted the author to leave the room by pointing to the author and then to the door. The author did not leave the room but did move farther away from Rose. During the second session, Rose's father began with easy questions and then proceeded to the more difficult questions. Figure 1 shows that Rose responded to all the questions using NVB.


During the third and fourth sessions, the author served as the novel conversational partner, and Rose's father remained in the room. All other aspects of each session were structured in the same way as during the first two sessions respectively. As indicated in Figure 1, Rose failed to respond with VSB, PVB, or NVB during session three, when asked 10 easy questions. Figure 1 also indicates that Rose produced no communication behavior during session four, when asked 10 questions beginning with easy content and proceeding to more the difficult content.

Two hypotheses were developed from the results of the BEA. First, it was hypothesized that Rose used SM behavior to avoid communicating with unfamiliar individuals. This was corroborated by information gained from the literature and by information in the parent report indicating that Rose speaks to other select individuals in person and on the phone. This hypothesis is also consistent with the findings reported by Schill et al. (1996). Therefore, habituation of a novel person was incorporated into the intervention. Habituation is defined as a decrease in response to a stimulus (in this case a person) after repeated presentations or exposure (Cooper et al., 2007).

The second hypothesis was that Rose's communication behavior was reinforced by the presence of her father, since she always produced some mode of communication when her father served as the social partner. As a result, the father was included in the design of the intervention.


Design. A single subject phase change design was implemented based on differential reinforcement of communicative behavior (DRC). Differential reinforcement entails reinforcing one response class while withholding reinforcement for another response class; it is a defining feature of shaping new behaviors (Cooper, Heron, & Heward, 2007). In this study, DRC began with differential reinforcement of NVB, then PVB, and finally audible VSB (Giddan et al., 1997). The criteria for differentially reinforcing the communicative behaviors are discussed in the following paragraphs. Phase changes occurred at each new level (NVB, PVB, VSB) from nonverbal behavior to audible speech.

Throughout the phase changes, natural withdrawals of the intervention were introduced by school holidays. This permitted the examination of treatment effect maintenance through quick probes of the dependent variables following each holiday break. In addition, a follow-up phase was conducted after the cessation of the intervention to examine maintenance and generalization of the dependent variable after the completion of the intervention.

Baseline. During baseline, the rate of each mode of communication was measured under a no demand situation. No questions were posed but the author and Rose were in the room together exploring the environment as the author commented about the room. Baseline took place in the same room where intervention would occur. Baseline data were collected over a one and half week time period and were used to provide a comparison with the demand phases.

Intervention. Differential reinforcement of modes of communication (DRC) was utilized to increase the frequency of NVB, then PVB, and then VSB in the presence of the researcher (an unfamiliar communication partner). The intervention phase began with the delivery of tangible reinforcement (preferred toys, candy) by the researcher following all attempts to communicate, and sessions were conducted two times a week. NVB was differentially reinforced first and continued until Rose began to use VSB (Giddan et al., 1997). Thus, the intervention phases shaped successive approximations through differential reinforcement of increasingly verbal communication modes until functional vocal communication was established (Schill et al, 1997; Miltenberger, 2007). The intervention phases spanned over a period of 13 weeks excluding time out for official school vacations. When vacations are included, the study extended over a four-and-a-half-month period.

Phase I: Reinforcement of NVB and PVB. During this phase, Rose was encouraged to speak as the researcher provided her with the following instructions: "Rose, you and I are going to get to know each other today. You may choose to respond in any way that you wish to respond." Following the instruction, the author asked conversational/ inquiry-based questions in an attempt to evoke communicative behavior For example, "How is your day?" and "Did you watch a cartoon this morning before school?" All NVB and PVB were immediately and differentially reinforced. Reinforcement consisted of briefly removing the demands of the conversational questions and allowing Rose to play with preferred toys in the presence of the author. This phase of intervention was continued until Rose produced an increase in both NVB and PVB. Once PVB occurred three times, NVB was no longer reinforced.

Phase II: Reinforcement of PVB. This phase marked the beginning of differential reinforcement for PVB. Rose was instructed that she would no longer receive access to reinforcement for using NVB but rather only when she used PVB. She was provided with examples of PVB (e.g., any mouth movement that resulted in sound) and with non-examples (e.g., gestures, pointing, and mouth movements without sound).

Withdrawal of Treatment: A four-day school holiday followed at the end of Phase II. This resulted in a natural withdrawal of treatment. A probing day took place following the break.

Phase III: Reinforcement of VSB with Parental Presence. The third phase of intervention occurred following the four-day withdrawal from intervention. PVBs had dropped in frequency and NVBs had increased. For each session within this phase, Rose was with her father inside of the room on her father's cell phone while a novel communication partner (the author) was outside of the room on another cell phone. As in Watson (1992), Rose was allowed to use the cell for responding vocally to questions asked by the novel communication partner. Prior to beginning this session, Rose's father was taught the process of DRC; when to deliver reinforcement; and when to withhold reinforcement. The author initiated the phone call at the beginning of the session. Within two-minutes of the phone call, Rose used VSB to communicate with the novel partner on the phone, and she immediately earned access to the toys in the classroom. Once VSBs occurred, only VSBs were reinforced and habituation to the novel person (the researcher) began.

Schill et al. (1996) described habituation as a stimulus fading procedure. In the present research, the novel communication partner gradually opened the door and moved inside the classroom to a distance of approximately six feet from Rose while Rose faced the other direction (Rye et al., 1999; Watson et al., 1992). Rose was aware that the researcher was getting closer to her, but she continued to look away and to keep her back to the researcher while looking at and remaining next to her father.

Withdrawal II: A second withdrawal occurred because of another school break (a three day break), which interrupted Phase III. A probe session was conducted following the break. The probe permitted evaluation of the generalization of Rose's communicative behavior after talking on the cell phone. Differential reinforcement of vocal behavior continued. However, vocalizations dropped to zero.

Phase III, continued: Phase III training continued until the end of the year.

Withdrawal with Parent Support: At the end of the school year, the family planned to leave town for the majority of the summer. To support the gains made by Rose during the school year, her father was provided with written instructions for providing DRC and with a treatment integrity checklist to self-record the implementation. He was asked to conduct three weekly sessions involving novel communication partners and to videotape one of them. He was also encouraged provide Rose with access to preferred items when she vocalized her wants and needs in the presence of novel communication partners.

Generalization and Maintenance. A follow-up observation was conducted at one month into Rose's first grade school year, about three months after the end of the intervention. At this time, it had been reported that Rose had begun speaking during class. To validate this, the author observed Rose in her classroom for 30 minutes on three separate days and recorded the frequency of the same variables targeted throughout the study (NVB, PVB, and VSB).


Treatment integrity was assessed during each session by using a Treatment Integrity Checklist (TIC) developed by the author and provided in Appendix A. This checklist outlined the procedures for implementing differential reinforcement of communicative behavior (DRC). Each item on the TIC was checked off as it occurred. One observer used the checklist to document the fidelity of the intervention during the session. Another observer used the checklist while observing the sessions on videotape. The results documented by the two independent observers were used to achieve inter-observer agreement (IOA) for treatment integrity. Results indicated 100% IOA for treatment integrity across all sessions



Based on initial assessment data collected by interviewing Rose's father, Rose had an unremarkable developmental history. On both the Eyberg Child Behavior Inventory and the Achenbach Child Behavior Checklist for parents (CBCL), Rose reportedly demonstrated age-appropriate behavior. The scores on the Social Skills Rating System (SSRS) indicated that Rose's social skills were age-appropriate and that her level of engaging in inappropriate behaviors was lower than same-age typical children. Overall, this information was consistent with the pre-treatment compliance questionnaire that revealed an above average compliant child.

Rose's teacher completed the Functional Assessment Informant Record for Teachers (FAIR-T), SSRS, CBCL for teachers, and Sutter-Eyberg Student Behavior Inventory - Revised (SESBI-R). Each of these rating skills was consistent with the father's reports regarding current social skills and compliance within the classroom.

Results from the preference assessment interview suggested that candy and access to preferred toys were the most reinforcing consequences for Rose and thus were used during intervention.

All results provided below are rates of behavior observed during a session.


During the first session of BEA Rose produced 2 NVBs, 0 PVBs, and 0 VSBs. During the second session Rose produced 10 NVBs but still zero PVBs and VSBs. During sessions three and four the author served as the novel conversational partner replicating the two sessions with her father. During phase three, Rose produced 2 NVBs but zero PVBs and zero VSBs. In the final phase, none of the communication behaviors (VNB, PVB, and VSB) were used. Figure 1 below illustrates the number of times per session that Rose responded to the ten questions posed using any of the communication behaviors.


The data collected during baseline yielded a low and descending trend for NVBs while PVBs and VSBs were not observed at all. Following three sessions of a downward trend with NVBs and a zero levels of PVBs and VSBs, the intervention was initiated.


Phase I consisted of DRC for NVBs. However, all communicative responses were reinforced if they occurred. In these sessions, NVBs increased, PVBs occurred during the third session, and VSBs remained at zero. When PVBs occurred at least three times, Phase II was initiated.


In Phase II, DRC targeted PVBs, and NVBs were faded systematically. Initially, NVBs increased but then dropped. PVBs initially remained stable but began to increase as the phase came to an end due to a three-day school recess.


Brief generalization probes were conducted when Rose returned from the three-day recess. Results showed a slight increase in PVBs and a slight decline in NVBs. Therefore, phase three was initiated.


Phase III began with DRC targeting PVBs. Additionally, a gradual habituation procedure was used to promote VSBs during conversation with a novel communication partner (the author). Initially, Rose was encouraged by her father to talk to the novel communication partner on a cell phone while the partner was in another room. Gradually, the novel communication partner moved into the room in closer proximity to Rose. During this procedure, both NVBs and PVBs decreased and VSBs appeared with a steady upward trend.


Phase III was followed by another school recess, which interrupted treatment sessions for one week. Upon return from recess, a brief probe session was conducted utilizing DRC of only PVBs and VSBs. The result was a reemergence of PVBs and NVBs with zero production of VSBs. To recapture the gains made prior to the recess, phase III was implemented again.


During this phase, all observations were made by the researcher in the classroom during instruction. Event recording was used to document the rate per minute of NVBs, PVBs, and VSBs during 30 periods across three days. Rate was estimated for each modality by dividing the total number of behaviors by the length of the observation interval. Results showed that all three modes of communication occurred, with PSBs occurring more frequently NVBs or PVBs

Rose's patterns of responding across baseline, intervention phases, withdrawals, and generalization is summarized on Figure 2, below.



Parent acceptability was rated using a modified version of Treatment Evaluation Inventory Short Form (TEI-SF) pre-and post-treatment. Questions were modified based on the specific intervention. The 9 items were scored on a 5 point Likert scale where 5 = strongly agree and 1 = strongly disagree. The pre-test score equaled 38 and the posttest equaled 41. Thus, the parent's pre-and post-treatment responses reflected similar high levels of acceptability, with post-test ratings appearing slightly than pre-test ratings.


This study employed a package of interventions to increase the vocal speech behavior of a 6-year-old kindergarten girl known here as "Rose." First, a BEA procedure based on Schill et al. (1996) was used to assess the function of Rose's SM behaviors, resulting in two hypotheses: (1) attention from novel individuals reduced Rose's vocal speech behavior, and (2) Rose's vocal speech behavior was reinforced by attention from her father. These hypotheses were used to develop the intervention. The reduction of Rose's VSB in the presence of novel conversational partners followed a pattern similar to the participant in Giddan et al.'s (1997) study. Therefore, the intervention for Rose was modeled after the intervention designed for Giddan et al.'s participant. It began by establishing rapport and differentially reinforcing all modes of communicative behavior, beginning with NVB. Because multiple weeks passed with no evidence of audible vocal behavior, a stimulus fading procedure was added based on the second hypothesis and adapted from previous work of others (Schill et al., 1996; Watson, 1992). During this phase of the intervention, audible vocalizations included an average of 77 words per session. Evidence of the success of this intervention package was further observed during a probe session after the completion of the intervention in an untrained context where Rose had not spoken previously. At this point, Rose used NVB to gain teacher attention and she produced VSB in front of her teachers and peers. These are true indications of her social development.

The use of a BEA aided the intervention by directly revealing hypotheses regarding the functions of Rose's SM behavior. The analysis allowed the researcher to intervene with confidence, and it provided a framework for designing the interventions. It is reasonable to assume that the BEA results led to an intervention strategy that was more valid and reliable than one based on anecdotal information alone.

Teachers, researchers, and therapists may consider the procedures and results described in this case study if they are faced with a similar participant who demonstrates an escape-maintained social-communication deficit such as SM. More generally, the results of this study support the potential value of using BEA results to design behavioral interventions for problem behaviors with topographies that deviates from the more typical overt behaviors often requiring remediation in a classroom setting.


Although the procedures designed for this study successfully increased the vocalizations of a kindergartener with SM and supported the findings of previous studies, generalization of these findings to other individuals with SM, especially if the SM is not escape-maintained, require caution. First, these procedures may be of limited use for verbal children who refrain from talking with any communication partners. In such cases, it may be more difficult to establish rapport, conduct a BEA, and rule out other communicative disorders that may be an underlying cause of the SM. Also, differential reinforcement of progressive vocal approximations could be more challenging and less effective if the behavior serves more than just an escape/avoidance function (e.g., if the behavior is multiply maintained by external variables, such as unintentional reinforcement by others).

Another limitation relates to the multiple components of the intervention. When multiple components are involved, it is difficult to determine which component of the intervention was most effective for improving a behavior. It is also difficult to know how the sequencing of components may have influenced the target behavior and it is difficult to discern which components were actually necessary to change the behavior and which may have been unnecessary.

In future research, it is recommended that the generalization phase of similar interventions be conducted in multiple settings and across multiple novel social partners.


Achenbach, T. M. (1991). Child Behavior Checklist/4/18 and 1991 Proflie.. Burlington, VT: University of Vermont, Department of Psychiatry.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC.

Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Applied behavior analysis: Second edition. New Jersey: Upper Saddle River.

Cunningham, C., Cataldo, M., Mallion, C., & Keyes, J. (1983). A review and controlled single case evaluation of behavioral approaches to the management of elective mutism. Child and Family Behavior Therapy, 5, 25-49.

Eyberg., Sheila. (1999). Eyberg Child Behavior Inventory. Odessa, FL: Psychological Assessment Resources.

Eyberg, S., & Pincus, D. (1999). Sutter-Eyberg Student Behavior Inventory-Revised. Odessa, FL: Psychological Assessment Resources.

Edwards, R. P. (2002). A tutorial for using the functional assessment informant record for teachers. Proven Practice: Prevention and Remediation Strategies for Schools, 4, 31-38.

Giddan, J. J., Ross, G. J., Sechler, L. L., & Becker, B. R. (1997). Selective mutism in elementary school: Multidisciplinary interventions. Language Speech, and Hearing Services in Services in Schools, 28, 127-133.

Gresham, F. M., & Elliot, S. N. (1990). Social Skills Rating System. Circle Pines, MN: American Guidance Service.

Gresham, F., Watson, T., & Skinner, C. (2001). Functional behavioral assessment: Principles, procedures, and future directions. School Psychology Review, 30 (2) pp. 156-172.

Iwata, B. A. (1994). Functional analysis methodology: Some closing comments. Journal of Applied Behavior Analysis, 27, 413-418.

Kazdin, A. E. (1982). Single case research designs: Methods for clinical and applied settings. New York, NY: Oxford University Press.

Kehle, T. J., Madaus, M. R., Baratta, V. S., & Bray, M. A. (1998). Augmented selfmodeling as a treatment for children with selective mutism. Journal of School Psychology, 36, 247-260.

Krysanski, V. (2003). A brief review of selective mutism literature. The Journal of Psychology, 137 (1), 29-40.

Leonard, H. & Dow, S. (1995). Selective mutism. Anxiety disorders in children and adolescents. New York: The Guilford Press.

Lyst, A. M., Gabriel, S., O'Shaughnessy, T. E., Meyers, J., & Meyers, B. (2005). Social validity: Perceptions of check and connect with early literacy support. Journal of School Psychology, 43(3), 197-218.

Manassis, K., Fung, D., Tannock, R., Sloman, L., Fiksenbaum, L., & McInnes, A. (2003). Characterizing selective mutism: Is it more than social anxiety? Depression and Anxiety, 18, 153-161.

Masten, W. G., Stacks, J. J., Caldwell-Colbert, A. T., & Jackson, J. S. (1996). Behavioral treatment of a selective mute Mexican-American boy. Psychology in the Schools, 33, 56-60.

Moore, J. W., Doggett, R. A., Edwards, R. P., & Olmi, D. J. (1999). Using functional assessment and teacher-implemented functional analysis outcomes to guide inter vention for two students with attention-deticit/hyperactivity disorqer. hfoven Practice, 2,3-9.

Nolan, J. D. & Pence, C. (1970). Operant conditioning principles in the treatment of a selectively mute child. Journal of Consulting and Clinical Psychology, 35, 265-268.

Richards, C. S.. & Hansen, M. K. (1978). A further demonstration of the efficacy of stimulus fading treatment of elective mutism. Journal of Behavior Therapy and Experimental Psychiatry, 9, 57-60.

Rye, M. S. & Ullman, D. (1999). The successful treatment of long-term selective mutism: A case study. Journal of Behavioral Therapy and Experimental Psychiatry,30, 313-323.

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.

Stein, M. (2001). Challenging case: Developmental delays and regression, selective mutism. Pediatrics, 107 (4), 926-929.

Stone, B., Kratochwil, T., Sladezcek, I. & Serlin, R. (2002). Treatment of selective mutism: A best evidence synthesis. School Psychology Quarterly, 17(2), 168-190.

Watson, T. S. & Kramer, J. J. (1992). Multimethod behavioral treatment of long-term Selective mutism. Psychology in the Schools, 29, 359-366.

Special thanks are extended to R.A. Doggett for mentoring this project and to Susan Elswick and Jenny Hayes for their assistance with data collection.


Laura Baylot Casey, Ph.D.

College of Education

The University of Memphis

405 Ball Hall

Memphis, TN 38152

Phone: (901) 678-1160

Laura Baylot Casey The University of Memphis
COPYRIGHT 2012 Behavior Analyst Online
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Casey, Laura Baylot
Publication:The Journal of Speech-Language Pathology and Applied Behavior Analysis
Article Type:Report
Date:Aug 1, 2012
Previous Article:Caregiver training in an AAC intervention for severe aphasia.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters