Minnesota State University, Mankato
Minnesota State University, Mankato
Cornerstone: A Collection of Scholarly
Cornerstone: A Collection of Scholarly
and Creative Works for Minnesota
and Creative Works for Minnesota
State University, Mankato
State University, Mankato
All Theses, Dissertations, and Other Capstone
Projects
Theses, Dissertations, and Other Capstone
Projects
2019
Using Behavioral Skills Training to Teach Children with Autism to
Using Behavioral Skills Training to Teach Children with Autism to
Seek Help from Law Enforcement Officers When Lost
Seek Help from Law Enforcement Officers When Lost
Greta Kos
Minnesota State University, Mankato
Follow this and additional works at: https://cornerstone.lib.mnsu.edu/etds
Part of the Clinical Psychology Commons, and the Educational Psychology Commons
Recommended Citation
Recommended Citation
Kos, G. (2019). Using behavioral skills training to teach children with autism to seek help from law
enforcement officers when lost [Master’s thesis, Minnesota State University, Mankato]. Cornerstone: A
Collection of Scholarly and Creative Works for Minnesota State University, Mankato.
https://cornerstone.lib.mnsu.edu/etds/932/
This Thesis is brought to you for free and open access by the Theses, Dissertations, and Other Capstone Projects
at Cornerstone: A Collection of Scholarly and Creative Works for Minnesota State University, Mankato. It has been
accepted for inclusion in All Theses, Dissertations, and Other Capstone Projects by an authorized administrator of
Cornerstone: A Collection of Scholarly and Creative Works for Minnesota State University, Mankato.
Using Behavioral Skills Training to Teach Children with Autism to Seek Help From Law
Enforcement Officers When Lost
By
Greta Kos
A Thesis Submitted in Partial Fulfillment of the
Requirements for the Degree of
Masters of Arts Degree
In
Clinical Psychology
Minnesota State University, Mankato
Mankato, Minnesota
Summer 2019
June 18, 2019
Using Behavioral Skills Training to Teach Children with Autism to Seek Help From Law
Enforcement Officers When Lost
Greta Kos
This thesis has been examined and approved by the following members of the student’s
committee.
________________________________
Advisor
Angelica Aguirre, PhD
________________________________
Committee Member
Daniel Houlihan, PhD
________________________________
Committee Member
John O’Neill, PhD
Abstract
Children with Autism Spectrum Disorder (ASD) are at an elevated risk of becoming lost in
situations due to running away or wandering behaviors. When children with ASD become lost it
may be difficult for them to effectively and efficiently seek help from community helpers in a
safe manner due to communication deficits and poor social skills. This study aims to teach
children with ASD how to identify when they are lost and how to seek help from police officers
in the community by vocalizing the problem and using a communication card. Generalization
probes were conducted with three children diagnosed with ASD before and after behavioral
skills training (BST) to determine mastery of help seeking behaviors. Remedial BST and in situ
trainings were used to increase generalization. Two of the three participants were able to
independently seek help from a confederate law enforcement officer and one participant required
a gestural prompt.
Keywords: BST, law enforcement, help seeking, lost identification
Table of Contents
Introduction………………………………………………………………..…….………………….1
Method…………………………………………………………………………………..…….….14
Results……………………………………………………………………………….………….….21
Discussion………………………………………………………………………….………….….23
References………………………………………………………………………….………….….28
Figure 1………………………………………………………………….………….………..……33
Table 1………………………………………………………………….…………………………34
Appendix C…………………………………………………………………….….………………35
1
Using Behavioral Skills Training to Teach Children with Autism to Seek Help From Law
Enforcement Officers When Lost
Introduction
The core symptoms of autism spectrum disorder (ASD) in children include
communication deficits, and poor social interactions with hyper- or hyporeactivity to sensory
inputs (American Psychiatric Association, 2013). Forty-nine percent of children with ASD
attempt to run away from a safe location at least once by the age of 4 (Anderson et al., 2012).
Running away can be very dangerous for the child and stressful for the family system.
Communication and social deficits may result in an inability to independently request or accept
assistance from community helpers, such as law enforcement, after running away from a safe
space. Behavioral analytic approaches have been used in the literature to teach children with
ASD to seek help from community members when lost in public (Bergstrom, Najdowski, &
Tarbox, 2012; Carlile, DeBar, Reeve, Reeve, & Meyer, 2018; Pan-Skadden et al., 2009; Taylor,
Hughes, Richard, Hoch, & Rodriquez Coello, 2004). Law enforcement officers are an important
first point of contact for children with ASD and typically developing children as they are trained
and trusted in crisis situations (Compton et al., 2014). For this reason, training needs to be done
to teach children with ASD to effectively respond to and communicate with community helpers,
such as law enforcement officers. The interactions that children with ASD have with law
enforcement officers is an important area of study moving forward (Archer & Hurley, 2013).
This study aims to use a behavioral analytic approach to teach children with ASD to
appropriately and effectively seek help from law enforcement officers when lost in public.
Interactions with Law Enforcement Officers for children with ASD
Recent reports indicate the rate of violent crime against people with disabilities is twice
as high compared to crimes against people with typical development, however, there continues to
2
be a gap in the literature on how to teach appropriate interactions involving children with autism
and law enforcement officers (Cheeley et al., 2012; Teagardin, Dixon, Smith, & Granpeesheh,
2012). A common parent concern is how a child with ASD will behave in an interaction with a
law enforcement officer and how the law enforcement officer will respond to the child’s
behavior. According to Rava, Shattuck, Rast, and Roux (2017), 20% of youth with ASD are
stopped by law enforcement by the age of 21. The social deficits many children with ASD
exhibit may come off as engaging in non-compliance towards law enforcement (Dogan et al.,
2017). Without proper officer and child training, these interactions have the likelihood to be poor
if the child is displaying emotional or behavioral outbursts and is unable to identify themselves
as an individual with ASD. Proper officer trainings could be those that lead to officers
identifying individuals with ASD and lead to understanding of effective communication with
individuals with ASD. These results should also be maintained over time and training refreshers
should be implemented in order to increase maintenance.
Although there are reports of officer trainings for working with people with special
needs, there have only been a few studies to show their benefit (Bailey, Barr, & Bunting, 2001;
Teagardin et al., 2012). These studies have shown an increase in awareness and identification of
individuals with special needs but failed to show maintenance of the trained skills (Bailey et al.,
2001; Teagardin et al., 2012). In another study conducted by Tint, Palucka, Bradley, Weiss, and
Lunsky (2017), parents of 284 adolescents and adults with ASD were asked to describe police
interactions with their child and their satisfaction with the police involvement over a 12-18-
month period. It was reported 16% of those sampled had an interaction with an officer. The
primary reason for the officer interaction was due to aggressive behavior. Part of the reason
individuals with ASD engage in aggressive behaviors can be attributed to their deficits in
3
communication (Carr & Durand, 1985). According to this study, individuals with ASD that have
interactions with officers are more likely to have a history of aggression, live outside the home
possibly receiving group home services, and have a parent with high caregiver strain.
Additionally, individuals with more police interactions are less likely to be able to afford
services and they are less likely to have structured daily activities. Police involvement was
reported to increase agitation in approximately 32% of interactions or have no effect in 17% of
interactions in the given study (Tint et al., 2017). This is a concerning statistic as officers should
have appropriate training to deescalate the situation. One possible way to combat this and the
fear of negative interactions with law enforcement officers is to reinforce children with ASD to
have positive interactions with law enforcement by teaching them who law enforcement officers
are, how officers can help them and giving the children a way to appropriately communicate
their diagnosis and their need for help. This could be done by examining the behavior analytic
literature on teaching safety and community skills.
Behavior analytic approach to teaching safety/community skills to individuals with
developmental disorders
Behavioral analytic approaches are commonly used to teach safety skills to individuals
with developmental disorders, particularly children with ASD. These skills include help-seeking
behaviors when lost (Bergstrom et al., 2012; Carlile et al., 2018; Hoch, Taylor, & Rodriguez,
2009; Pan-Skadden et al., 2009; Purrazella & Mechling, 2013; Taber, Alberto, Seltzer, &
Hughes, 2003; Taylor et al., 2004), preventing gun play (Himle, Miltenberger, Flessner, &
Gatheridge, 2004), enhancing fire safety (Garcia, Dukes, Brady, Scott, & Wilson, 2016) and
teaching abduction prevention (Bergstrom, Najdowski, & Tarbox, 2014; Gunby & Rapp, 2014;
4
Ledbetter-Cho et al., 2016). Training safety skills using a behavioral analytic approach
commonly utilizes a number of different procedures.
For example, Rodriguez, Levesque, Cohrs, and Niemeier (2017) used an interrupted
chain procedure in a multiple probe design to teach three children with ASD or pervasive
developmental disorder to request help from others with difficult tasks. Preassessments were
conducted to determine fine motor capabilities similar to the behavioral chains used later in the
general procedures (e.g., operating scissors). Pre-test probes were conducted once or twice per
clinic appointment with a total of 12 trials per session with 6 being capable trials and 6 being
incapable trials. During incapable trials, an interrupted chain procedure was used to manipulate
the materials so one link of the behavioral chain could not be completed without assistance,
capable trials were identical to incapable trials but all parts of the behavior chain could be
completed without assistance. During probe sessions, help was given upon request but no other
reinforcements were offered. Teaching began with a 5s prompt delay to attempt the task before a
partial physical prompt and a 0s prompt delay to request help after attempting the incapable task,
this was then faded to a 5s prompt delay to attempt and a 2s prompt delay to help. Generalization
probes were then completed similar to pre-test probes. Correct responding increased for
participants across phases with 100% correct responding consistently observed. Expanding on
requesting help with difficult tasks, behavioral analytic approaches have also been used to teach
help seeking when lost.
Teaching help seeking behaviors when lost has been done in numerous ways using rules,
technology alone or technology with the addition of a communication card. Bergstrom and
colleague (2012) used a multiple baseline design to teach 3 boys with ASD to seek help when
lost in public. Prior to each teaching trial, the participants were given rules on what to do if they
5
were lost, including yelling for their parents, finding a worker (i.e., store cashier) and telling the
worker that they are lost. During training, the participants were provided verbal praise contingent
on a correct resonse (e.g., asking for help) or provided verbal prompts if they engaged in an
incorrect response. Researchers implemented a 0-3 rating scale to determine mastery levels of
following the steps. Participants received 1 point for engaging in each of the behaviors All three
participants met mastery levels in post-test training by receiving consistent scores of 3 across
different settings.
Taylor et al. (2004) implemented technology using a pager in teaching help-seeking
behavior using a multiple baseline probe design across participants with three teenagers
diagnosed with ASD in a community setting. Participants were paged in the community setting
and instructed to hand a communication card to a community member indicating that they were
lost and needed help. The pager was used as a conditioned stimulus that could be generalized
across settings. Responses stabilized at the 100% correct responding mark around 55 sessions
after training for participant 1, 20 sessions after training for participant 2 and 10 sessions after
training for participant 3.
Similarly, Hoch et al. (2009) used the implementation of a cell phone instead of a pager.
This study used a multiple baseline probe design across three teenage males diagnosed with
ASD. During probes, when participants were separated from caregivers they received a phone
call and were expected to answer the call, find an adult, state they were lost, exchange a
communication card, and wait with the adult until they were reunited with the caregiver. The
community members were blind to the study and data was also collected on their responses.
Participants were taught how to answer their phone and reinforced for answering the phone call
and correct responses. Participants had additional training in the community and practiced
6
approaching outside community members who were blind to the study and asked for help. All
participants were able to answer their cell phone, find a community member, ask for help and
wait with them during post-test community probes.
In addition, Purrazella and Mechling (2013) used video modeling on seeking help when
lost using a multiple probe design across participants. They taught three individuals aged 24-29
with moderate ID to take and send video captions of their location when they were lost in the
community. The intervention consisted of a task analysis using video modeling, picture prompts,
and instructor prompts from turning on the phone to sending the video. Teaching was conducted
in an isolated classroom setting at the community college that all participants attended. Probe
and generalization probe sessions took place in the community at three separate locations
including a church, a residential setting and in a retail area. The participant and the researcher
road public transportation for a set amount of time and then got off and walked two blocks
before arriving at the location where the probe would take place. It was found that all three
participants learned to follow a sequence of steps for operating the iPhone 4 with video capturing
and to use the skill across three unfamiliar community settings.
Continuing with cell phone use, Taber et al. (2003) conducted a study with 6 secondary
students aged 14-18 with moderate cognitive disabilities who were unable to identify when they
were lost or unable to dial a cell phone to call for assistance. This study used a multiple probe
design across participants. All methods were similar using a task analysis but participants were
divided into “Group A” or “Group B.” One group was taught to answer a phone call when they
were lost and provide a detailed description of their location. The other group was taught to dial
a number from speed dial on the phone to call for assistance and provide a detailed description of
7
their location. Results indicated that all students learned to either answer a ringing cell phone or
use the speed dial function to call for assistance.
Research has advanced to use more modern technology such as FaceTime that can be
seen in a study conducted by Carlile et al. (2018), who implemented a package intervention to
teach children with ASD to seek help when they were lost. The researchers implemented a high
technology response situation, FaceTime, and low technology response situation, a
communication card. The researchers used a multiple-baseline across participants design
imbedded in an alternating treatment design with the two technology situations to evaluate the
effects of video modeling and programming common stimuli to teach the low and high-tech help
seeking responses. Participants were six males ages 3-14 diagnosed with ASD. Each participant
was trained in both the high tech and low tech condition of help seeking. The intervention used
video modeling to show each participant how to respond if they became lost in public. After the
video modeling, the participant was then placed in a mock situation inside the classroom and
praised for correct responding and interrupted with the video modeling re-presented for incorrect
responding. During probes, participants were placed in two groups. “Group A” was expected to
make a FaceTime call and then approach a store worker while “Group B” was expected to
answer a FaceTime call and then approach a store worker. Participants in this study drastically
increased their help seeking responses when lost improving from 0% correct to 100% correct and
they maintained the behaviors at follow-up and during community probes. Another common
behavior analytic method to teaching safety skills is behavioral skills training (BST).
Studies using BST to teach safety and community skills
BST has been used to teach a variety of behaviors to individuals with and without
intellectual disabilities for numerous years (Aguirre, O’Neill, Rehfeldt, & Boyer, 2014). BST
8
uses instruction, modeling, rehearsal and feedback in a variety of simulated situations to teach a
skill or skills to an individual or a group of individuals. More specifically, instruction includes
teaching participants to label, identify and provide rationale for engaging in the behavior.
Modeling includes describing and demonstrating the behavior. Rehearsal and feedback involves
practice, feedback and delivery of a consequence (Aguirre et al., 2014). These fundamental
components make up BST and its effectiveness. According to Miltenberger 2008, BST is more
effective than an informational approach without an active learning component. Numerous
studies have used BST to teach safety skills to typically developing children regarding gunplay
(Himle et al., 2004; Jostad, Miltenberger, Kelso, & Knudson, 2008; Miltenberger et al., 2004;
Miltenberger et al., 2005), to teach fire safety to children with ASD (Garcia et al., 2016), to
respond to abduction lures for children with ASD (Bergstrom et al., 2014; Gunby, Carr, &
Leblanc, 2010; Gunby & Rapp, 2014; Ledbetter-Cho et al., 2016), to respond to abduction lures
for typically developing children (Johnson et al., 2006) and to teach help-seeking when lost to
typically developing children (Pan-Skadden et al., 2009).
BST has been a primary tool used to teach firearm safety to typically developing children
over recent years. Himle et al. (2004) started with 10, four and five year old typically developing
children who were taught gun safety. This study implemented a multiple baseline design across
participants using BST, remedial trainings and in situ trainings when needed for most
participants to reach consistent satisfactory scores in this study. Himle et al. scored each probe
on a scale of 0-3 based on their response to seeing the firearm in the room. Almost all
participants required the remedial training sessions and over half required in situ training. The
same study was replicated with six, 6- and 7-year-olds (Miltenberger et al., 2004). The
replication found similar results in that in situ training was still necessary for over half of the
9
participants even though they were an older developmental age. The study was adapted again
with 10, 4 and 5 year old children who received two BST sessions and 1 in situ training session
planned ahead of time (Miltenberger et al., 2005). This supported using both BST and in situ
training together for this developmental age.
In a similar study, Jostad et al. (2008) found that peer tutoring can be an effective way to
train children about firearm safety in order to train more children using less resources. Jostad et
al. used BST to train a group of 6- and 7-year-olds on firearm safety and then instructed them
how to train 4 and 5 year olds on firearm safety using BST. The researchers recorded the older
group training the younger group and found good results for the younger group meeting mastery
criterion (i.e., scores of 3). In a review of past studies, Miltenberger (2008) not only concluded
that BST was more effective than an informational approach to training, but, that an in situ
assessment is the only way to determine if the child will use the skill and that in situ training is
the most reliable method for producing generalization. Gunplay is a well known exemplar and its
components have also been shown in other areas like fire safety.
BST has also been used to teach fire safety. Garcia et al. (2016) implemented BST in a
nonconcurrent multiple baseline to teach fire safety to three, 4- and 5-year-old children
diagnosed with ASD. The training consisted of teaching the participants what to do when they
hear a fire alarm using modeling and rehearsal. Participants were taught to evacuate when they
heard a fire alarm and immediately find an adult to tell after evacuating. During generalization,
data collected showed an increase of skills implemented correctly. Generalization probes after 5-
weeks showed continued implementation of skills as well.
Teaching children with ASD to respond to abduction lures is another area where
numerous studies have been conducted using BST. Johnson et al. (2006) used BST with 50
10
typically developing 6- and 7-year-old children to teach them to say no, immediately walk away,
and tell an adult when they are presented with an abduction lure. This study used a 0-4 scale
regarding the response of the participant to the lure, 0 = agrees to leave with the abductor; 1 =
does not agree to leave with the abductor but fails to say ‘‘no,’’ get away, or tell an adult; 2 =
says ‘‘no’’ but does not leave the area or tell an adult; 3 = says ‘‘no’’ and leaves the area but
does not tell an adult; 4 = says ‘‘no,’’ leaves the area, and tells an adult. Participants were split
into three conditions; BST only, BST plus in situ training, or a control group. The study found
significant results for the BST and BST plus in situ training groups over the control group in
posttest trials.
Gunby and others (2010) taught abduction-prevention skills using BST with in situ
feedback to three children with ASD. The participants were taught to say no to the abduction
lure, get away from the situation and immediately tell an adult. All participants acquired the
skills and were maintained at a one month follow up. Gunby and Rapp (2014) extended the
previous work and used abduction lures after a high-probability request sequence simulating the
grooming or recruitment (i.e., priming) process. This included a wider range of abduction lures.
All three participants diagnosed with ASD acquired the safety response and maintained after the
one month follow-up. Bergstrom et al. (2014) extended the research further using primarily male
confederates due to the statistics on male versus female perpetrators. This study used a
nonconcurrent multiple baseline across three boys diagnosed with ASD. All participants in the
study learned to correctly respond to the abduction lures similar the previous study. This showed
support for generalization across environments and abductors. Recently, Ledbetter-Cho et al.
(2016) provided a further extension by including multiple lure types in the baseline and assessing
for maintenance of the target response across lures. This study was conducted with four children
11
diagnosed with ASD using a concurrent multiple baseline design. Improvements were found
from the baseline that generalized to novel settings and confederates. Additionally, these
improvements maintained at the four week follow-up.
The same concept of BST to teach safety responses to children with ASD can be used
when they are lost. Pan-Skadden et al., (2009) demonstrated this technique with typically
developing children. Three children, age 4 or 5 with no medical or developmental disabilities
participated in the study. The researchers used BST and in situ to teach the children help-seeking
responses if they become lost. The study implemented a 0-3 scale with 0 = participant remained
in the same location for 20 seconds after realization that they were separated from caregiver; 1
= demonstrated any behavior other than going up to the cashier; 2 = went to the cashier within
20 seconds and told the cashier that they were lost but did not give any personal information; 3
= went to the cashier within 20 seconds, stated their personal information, and stated that they
were lost. The experimenter instructed the participants on what to do if they are lost in a store,
modeled the behavior with dolls or acting, rehearsed with the participants and gave feedback on
their rehearsals. Thirty minutes following the BST, the caregiver went to “shop” with the
participant. At the store, an assessment occurred in which the caregiver “lost” the participant. If
the participant performed the steps in 20 seconds they received praise, if they did not perform the
steps in situ training was conducted. All participants were able to increase to a level 3 criterion
and maintain at the 2-week follow-up.
Statement of the Problem
Research has been conducted teaching children with ASD to seek help when they are lost
(Bergstrom et al., 2014; Carlile et al., 2018; Hoch et al., 2009; Taylor et al., 2004). As mentioned
previously, a BST model has been shown to be effective in teaching typically developing
12
children to seek help when lost (Pan-Skadden et al., 2009). However, there is a lack of literature
using BST to teach help seeking behaviors when lost to children with ASD even though there is
clear support for the effectiveness of BST for teaching safety skills to children with ASD
(Bergstrom et al., 2014; Garcia et al., 2016; Gunby & Rapp, 2014; Gunby et al., 2010; Ledbetter-
Cho et al., 2016). In addition, further research is needed to ensure generalization of help seeking
skills to community members such as law enforcement officers since they are the historically the
first point of contact in crisis situations. Using prevention of firearm injury as an exemplar,
Miltenberger (2008) concluded that the BST approach is more effective than an informational
approach to training safety skills.
Further, children with ASD are a vulnerable population who could be easily
misunderstood by authority figures for example, there are reports of law enforcement officers
harming children with ASD because they do not understand them (Associated Press, 2017). As
previously mentioned, approximately 49% of children diagnosed with ASD run away from a safe
place by the age of 4 (Anderson et al., 2012). When the child has ran away from a safe place for
any reason, it would be helpful for the child to be able to identify an officer in the community
and seek help in a positive manner. Tint and colleagues reported that 16% of adolescents and
adults diagnosed with ASD had an encounter with an officer, primarily due to an aggressive
behavior. Communication deficits sometimes can serve as the function of aggression and other
disruptive behaviors (Carr & Durand, 1985). These communication deficits and subsequent
disruptive behaviors may lead to more police encounters. By teaching children to seek help from
officers when lost, the percentage of interactions may increase in a positive and beneficial
manner for both parties. Parents of children with ASD show concern for interactions between
officers and their children (Rava et al., 2017). Further research is needed to teach children to
13
seek help from community helpers, like law enforcement officers, when they are lost in the
community to further extend the literature using a BST procedure. The purpose of this study is to
teach children with ASD to identify when they are lost and effectively and efficiently seek help
from a police officer in the community to enhance their safety skills.
14
Method
Participants and Settings
Three children with autism spectrum disorder aged 5-10 participated in this study. Adam
age 5, Liam age 6, and Ellie age 10. All participants were diagnosed by a mental health
professional according to the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition
(American Psychiatric Association, 2013). Participants were recruited from a local autism center
in Southern Minnesota. The Vineland was administered to all participants to assess their
communication, daily living and community skills. According to age, Adam was in the low
range for daily living, socialization and motor skills and in the moderately low range for
communication. Adam showed receptive language as a strength (e.g., following instructions,
engage in conditional discriminations). Liam was in the low range for communication, and
moderately low for daily living, socialization and daily living skills. Liam showed community
skills as a weakness. Ellie was in the low range across all areas; communication, daily living,
socialization and daily living. She showed interpersonal relationships as a strength. The Verbal
Behavior Milestone Assessment and Placement Program (VB-MAPP) was also administered to
assess participants’ receptive and expressive language repertoires. To be included in the study,
participants were expected to be in the higher level two and show some emerging level three
scores. Adam’s VB-MAPP scores indicated that he is able to request most of his wants and needs
with 3-6-word phrases, can label and identify most common items and functions of objects. He is
also able to answer and ask basic social and personal questions. Liam had very similar scores to
Adam but he also showed some emerging basic reading and writing skills for a 5-year-old. Ellie
can communicate independently with an Augmentative and Alternative Communication Device
(AAC), which is an application on an iPad. Ellie’s scores were in the high level 1 and emerging
15
level 2 indicating that she is able to identify and label basic common items and answer some
personal questions through her device. However, she requires some gestural prompting to tact
and engage in listener responding in two or more-word phrases. She also showed some emerging
echoics due to her only being able to imitate one syllable sounds. The parents of these children
also must have indicated a concern for their child’s safety and interactions with police officers.
Training and generalization probes took place at the local autism center in Southern Minnesota.
Trainings were conducted in cubicles with one table, two chairs and a small play area.
Generalization probes were conducted in an open gym with all distractors put away or a small
room with a two-way mirror, TV and gaming system which participants did not have access.
Materials
The materials for this experiment were identification cards for the participants. The
identification cards included the participants name, age, diagnosis, address, best form of
communication and emergency contacts (2 x 3 card). Visual stimuli of police officers (4 x 6
picture cards) were used during identification training along with distractor stimuli of other
community officers (e.g., fireman and doctor). Confederate officers were trained Research
Assistants dressed as officers.
Dependent Measure(s)
The primary dependent measure was a scale of 0-3 indicating the level of the participants
interaction with the police officer. 0 = the participant had no interaction with the law
enforcement officer within a minute of identification of a problem or did not identify that they
were lost, 1 = the participant approached the officer within 1 minute of identifying the problem
and stood less than 3 feet away but did not hand the identification card, 2 = the participant
approached the officer within 1 minute of identifying the problem, was less than 3 feet away and
16
handed the identification card to the officer, 3 = the participant approached the officer within 1
minute of identifying the problem, was less than 3 feet away, handed the identification card to
the officer and vocalized (through oral communication or with AAC) the problem. The secondary
dependent measure was the percentage correct of responding during identification training before
moving onto behavioral skills training.
Experimental Design
A non-concurrent multiple baseline across participants design was used. Pretest and
posttest generalization probes were conducted to probe the participants interactions with the
confederate officers. Behavioral skills training (BST) was the primary independent variable in
this study. BST began with identification training then moved into traditional components
including instruction, modeling, rehearsal and feedback. Identification training was implemented
to teach participants to discriminate police officers from other community helpers and identify
their roles following a similar format to traditional community helper teaching in Applied
Behavior Analysis programming. Instruction included verbal instruction regarding how to
approach an officer when lost. Modeling was done by the researchers to demonstrate identifying
when they are lost and approaching an officer for help. Rehearsal consisted of the researcher and
the child practicing interacting with law enforcement officers and feedback was given throughout
this stage. A Remedial BST and in situ trainings were used if the participant was not receiving
scores of 3 during posttest probes.
Interobserver Agreement (IOA)
Total IOA was conducted for 100% of probe sessions for Adam with a 91% agreement
and 88% of training sessions for 100% agreement. Total IOA was conducted for 100% of probe
sessions for Liam with a 100% agreement and 71% of training sessions for 100% agreement.
17
Total IOA was conducted for 100% of probe sessions for Ellie with a 97% agreement and 64%
of training sessions for 100% agreement. An agreement was defined as the two observers
recording the same response on the 0 to 3 scale or circling the same response on the data sheets.
A disagreement was defined as the two observers recording different responses on the data
sheets. Total IOA was calculated by dividing the number of agreements by the number of
agreements plus disagreements, and multiplying the resulting quotient by 100%.
Procedural Integrity
Procedural integrity (PI) was collected using a checklist designed specifically for the
study. Examples of items in the checklist include officer placement, officer communication,
participant attention and verbal praise delivered by experimenter during instruction. PI was
collected for 86% of probes and 63% of trainings for Adam for 100%. PI was collected for 86%
of probes and 50% of training for Liam for 100%. PI was collected for 91% of probes and 57%
of training with Ellie for 100%.
Procedures
Pretest generalization probes. Pretest generalization probes were conducted with the
participants in the local autism center. During pretest and posttest generalization probes,
confederates dressed as officers were placed in the autism center to portray real life scenarios.
The participants were put in situations of being lost and observers measured their interactions
with the confederate police officer based on the 0-3 scale. The identification card was placed in
the pocket or hand of the participant. Participants were not instructed on the use of the card
during pretest probes. During pretest and posttest generalization probes, the child and caregiver
walked together to a specific destination. The caregiver gave a directive task to the child to
complete when they were within 20 yards of the officer, such as “go grab that ball.” The
18
caregiver then slipped out of the room when the child began the task and was not looking at the
caregiver. Unknown to the participants, confederates were observing them to ensure safety,
including a confederate officer and 1-2 research assistants. The participant then had 3 minutes
from the time of the directive to seek help from the officer. If they did not perform the target
behavior in 3 minutes the caregiver returned. If there were any points of distress for the
participant, the confederate officer or experimenter was able to end the trial. During pretest
probes, if the participant walked up the confederate police officer or a community member
during session would end without reinforcement and the caregiver would return. If the
participant asked the caregiver where they were they were to respond with a neutral statement
(e.g., “I had to grab my keys”).
Behavioral skills training. The first stage of BST was identification, participants were
required to reach 100% mastery in each phase. Participants were taught how to identify a police
officer and how to discriminate officers from civilians. Listener training was used by showing
participants pictures of police officers and other professionals in a field of 4. Researchers
instructed participants to “point to the police officer.” Participants were required to point or
select all 3 different officers in order to move onto the next phase (100% across three 3-trial
blocks). If the participant answered incorrectly the researcher provided a least to most physical
prompt for the correct response followed by error correction then moved on to the next trial. Tact
training was implemented by showing pictures of 3 different police officers to participants and
asking “Who is this?” Participants were required to tact all three police officers in order to move
onto the next phase (100% across three 3-trial blocks). Intraverbal training consisted of asking
participants a series of questions about the role of police officers, including “who do we go to if
we need help?”, “what situations may we need help in?” and “who is the first person you look for
19
if you cannot find your parents?” Participants were required to respond correctly to each question
to move on to the next phase (i.e., 100% across three 3-trial blocks). Incorrect responding or no
response for both tact and intraverbal training were followed by an echoic prompt (i.e., “say
police officer”) then an error correction trial before moving on to the next trial. During listener,
tact, and intraverbal training, maintenance questions (e.g., motor imitation, listener responding of
motor actions) for each participant was interspersed on a VR 2 schedule. Reinforcement was
given following each correct response and included verbal praise and access to a tangible item
determined through a multiple stimulus without replacement preference assessment and paired
choice assessment prior to each session.
After the identification phase, verbal instruction using a visual prompt in the form of a
story was used to instruct the participants to give their identification card to police officers and
vocalize that they need help (see Appendix A). Rationale during instruction was provided in
order for participants to discriminate when they need help versus when they do not.
Phase two of training was modeling of the behavior using role play. Researchers
demonstrated lost scenarios and modeled how to approach an officer, how to hand their
identification card and how to ask for help. Modeling included researchers pretending to search
for their parents and stating “I am lost because I cannot find my mom, dad or therapist” then
walking through the steps of approaching a law enforcement officer and requesting help. Phase
three of training included rehearsal and feedback. It was required that during rehearsal the
participant received a rating of three on two consecutive trials before going into posttest
generalization probes. Participants were asked to pretend they were lost and practiced asking a
pretend officer for help in a controlled setting. Specific feedback was given to participants
whether they were correct or incorrect. If the participant did not meet mastery for two
20
consecutive posttest generalization probes they cycled back to rehearsal and feedback for
remedial BST before trying again.
Posttest generalization probes. After meeting the mastery criterion during rehearsal,
posttest generalization probes were conducted. Direct observation was used and participants
were given a score as indicated in the dependent measures section. If the participant received a
score of 3 during posttest generalization probes, they were reinforced through verbal praise. If
the participant asked a confederate community member (i.e., another adult) for help, they were
differentially reinforced and redirected to the officer (e.g. “oh ok, let’s look for an officer”). If
the participant did not receive a score of 3 for 2 consecutive posttest probes, remedial BST
instruction was implemented. If mastery was still not met on the subsequent posttest
generalization probe, the researcher used in situ training. In situ training included least to most
verbal and physical prompting during the generalization probe starting with verbal instruction
and going to physical prompting if needed.
Rule and consequence. A rule and consequence was added in addition to in situ training
for Liam and Ellie. This included the participant being told before each probe “if you tell the
officer that you are lost, then we can play with a toy.” A preferred item was selected from a
paired choice preference assessment before each probe. Following the rule, the probe would
begin with the therapist giving a directive to the child and if the child completed the steps they
would have access to the reinforcing toy. A time-prompt delay procedure an addition to the rule
and consequence was added for Ellie only.
Follow-up. Follow-up generalization probes were conducted to measure maintenance.
These probes were set up the same as pretest generalization probes and were done at two weeks
after the final posttest generalization probe for Adam and after three weeks for Liam
