According to the Suzy Lamplugh Trust, personal safety is defined as “an individual’s ability to go about their everyday life free from the threat or fear of psychological, emotional or physical harm from others”(Suzy Lamplugh Trust, n.d.). While ‘threat of harm’ is something many individuals are free from in their regular day-to-day, certain groups of people are more likely to experience risks to personal safety. For example, Adults with disabilities are 1.5 times more likely to be a victim of violence (Jones et al., 2012). Additionally, children are more likely to be exposed to violence than adults, and children with disabilities are 3.44 times more likely to be a victim of abuse than their peers without disabilities (“Facts about Children and Violence”, 2017; Davis, 2011).

Abduction is one type of threat to personal safety, particularly when it comes to the personal safety of children and children with disabilities. While most child abductions are committed by family members, child abductions by strangers are more dangerous and often result in more serious consequences such as abuse or even death (Kurt & Kutlu, 2019). According to the National Center for Missing and Exploited Children (NCMEC), over 1,300 cases (known to the NCMEC) of attempted or successful abduction occurred between January 1st, 2019 and December 31st, 2019 (The National Center for Missing and Exploited Children, n.d.).

While there are not many statistics available on what percentage of abductions involved a child with a disability, the NCMEC has published data on children with disabilities who have gone missing. According to the NCMEC, 952 children who were reported to have autism or Asperger syndrome as a medical condition, a mental health condition, or a special needs condition were reported missing between January 1st, 2007 and December 31st, 2016(Missing Children on the Autism Spectrum, 2016). The NMEC report did not list the reasons the children went missing; however, most missing cases were likely caused by wandering. While this paper does not directly address wandering, it should be noted that according to Ledbetter-Cho (2016) states that 50% of children elope from safe places which could potentially increase abduction risk.

While less at-risk than children, adults with disabilities are still 1.5 times more likely to be victims of violence than non-disabled adults. The likelihood of experiencing violence varies based on different types of disabilities (Jones et al., 2012). Jones et al. found that adults with intellectual disabilities were 1.6 times more likely and adults with mental illness were 3.86 times more likely to experience violence than non-disabled adults.

As with any survey data on controversial topics, the results from the above study may not capture the full extent of exposure to violence for individuals with disabilities. Finding statistics on violence towards individuals with disabilities has proven to be challenging. According to Sullivan (2002), children with disabilities are typically not reported in any crime statistic systems in the U.S., making it difficult to determine the risk of abuse for these children. Jones et al. (2012) conducted a meta-analysis of 17 studies to find pooled estimates of the prevalence and risk of violence against children with disabilities. Findings from Jones et al. (2012) indicated that one in four children with a disability will be a victim of violence. Of those children, 20.4 % will be a victim of physical violence and 13.7% will be a victim of sexual violence. Results of the meta-analysis also showed that children with intellectual or psychological disabilities are more likely to be victims of sexual assault that individuals with physical disabilities.

Many factors may contribute to why children with disabilities have an increased risk of victimization. According to Davis (2011), a potential reason that persons with disabilities may be more vulnerable to abuse than others is due to society’s response to the abuse. Davis (2011) states that adults may be less likely to report the abuse due to the individual’s disability status, making it more likely for abuse to continue. Parents may not report abuse occurring at a group home due to fear their child may not have other options for living arrangements, or parents or caregivers may not report abuse due to fear of legal ramifications.

For individuals with disabilities, the person being abused might not report the abuse due to not understanding what abuse is, or due to communication deficits (Davis, 2011). Watson (1984) suggests that deficits in social skills and judgment may be the reason for increased vulnerability to abuse. Additionally, communication deficits such as the inability to report abuse or seek help may increase risk (Lang & Frenzel, 1988). Lack of knowledge on how to defend against abuse and lack of education regarding appropriate sexual behavior has also been cited as potential reasons for increased risk (Sobsey & Varnhagen, 1988).

Individuals with disabilities are more likely to face risks to their personal safety such as those of abduction and abuse. This paper will cover the research in the field of behavior analysis on promoting personal safety behaviors, with a focus on individuals with developmental disabilities. Much research has been done within the field of behavior analysis on increasing the personal safety of individuals with disabilities, however there are still areas which warrant further research (Godish, Miltenberger, & Sanchez, 2017). This paper will analyze the current body of literature on personal safety in a hope to evaluate the progress made, gaps in the current body of research, and suggestions for future research.

The first section of this paper will cover the historical overview of teaching personal safety to children and adults with developmental disabilities. The second section of this paper will focus on the topic of personal safety skills as it relates to the seven dimensions of applied behavior analysis. After going over the background of this topic, the third section of the paper will review some of the applications regarding promoting personal safety skills. The fourth part of this paper will discuss ethical guidelines and decision making relating to teaching personal safety skills. The final part of this paper will discuss future directions regarding personal safety for individuals with disabilities. For this paper, the term personal safety will mainly be used to refer to the prevention of abuse and abduction.

Historical Overview

Some of the earliest prevention applications originated when the field of applied behavior analysis was instead referred to as behavior modification. According to Miltenberger (2008), behavior modification procedures have been applied to many areas to help people change a vast array of problem behaviors; one such application is ‘Prevention’. Behavior modification and now applied behavior analysis have been used to address many societal problems through teaching preventative skills. Physical abuse, sexual abuse, abduction, and neglect are some prevention areas that have been addressed in the behavior modification and behavior analysis literature (Miltenberger, 2008).

Behavior analytic preventative approaches to abduction or abuse before the 1980s are scarce (Poche, Brouwer, and Swearinger, 1981). Research on child abuse has historically focused on the characteristics of child molesters and treatment of sexual deviance after the fact (Forgione, 1976; Groth, 1980; Groth & Birnbaum, 1978; Panton, 1978). One of the first empirical studies published on the topic of abduction prevention was Poche, Brouwer, and Swearinger (1981). In this study, preschool children were taught how to respond to lures from strangers using modeling, behavior rehearsal, and reinforcement. Poche, Brouwer, and Swearinger (1981) demonstrated that through modeling, rehearsal, and reinforcement that preschool children could be taught to appropriately respond to lures for strangers.

Several years after Poche, Brouwer, and Swearinger (1981), Haseltine and Miltenberger (1990) demonstrated that adults diagnosed with mental retardation were able to learn and demonstrate abduction prevention skills. Haseltine and Miltenberger (1990) used behavioral skills training to teach abduction prevention skills to eight adults diagnosed with mild mental retardation. Results of their study found that seven of the eight participants learned the safety skills and maintained them at 6-month follow up. Additionally, a side-effects questionnaire found no evidence of emotional or behavioral side effects, which was one of the major concerns of conducting research in the area of abuse and abduction prevention. The positive outcomes of this study and previous abduction prevention studies opened the doors for more applied research to be done in the area of teaching personal safety skills to individuals with disabilities.

Haseltine and Miltenberger (1990) pointed out the need for future research to address the difficulties in assessing sexual abuse prevention skills. However, studies evaluating the effectiveness of programs for preventing sexual abuse of persons with developmental disabilities did not come about until 1998. Lumley and Miltenberger (1997) addressed the need for research in this area and outlined the necessary skills for preventing sexual abuse as well as how to assess the effectiveness such a program. The suggestions made for research in this area were based on previous research done in the area of abduction prevention and sexual abuse prevention studies done with children (Lumley & Miltenberger, 1997). Following the suggestions laid out in Lumley and Miltenberger (1997) Lumley et al. (1998) evaluated the effectiveness of behavior skills training on sexual abuse prevention skills of women diagnosed with mental retardation. This study incorporated naturalistic, in situ assessments in which participants were debriefed after. Most of the behavior-analytic applications that followed Lumley et al. (1998) focused on the prevention of sexual abuse. There continues to be very little behavior analytic research on preventing physical abuse of persons with disabilities (Doughty & Kane, 2010).

Many studies focusing on prevention of abuse or abduction utilize behavioral skills training, or most of its components, in order to teach safety skills to study participants (Lumley et al., 1998; Miltenberger & Thiesse-Duffy, 1988; Wurtele et al., 1986). Behavioral skills training involves providing instructions, modeling, rehearsal, and feedback for the behavior being taught. The use of behavioral Skills training emerged in the late 1960s and early 1970s, although it is unclear where the combination of instruction, modeling, rehearsal, and feedback first originated (Erhard, Falcomata, & Harmon, 2019). Some of the early applications involved variations of these main components but were not termed behavioral skills training. Early studies on these components have demonstrated their effectiveness in teaching people social skills. Braukmann et al. (1974) demonstrated that the use of instructions, rationale, demonstration, behavior practice, and feedback was effective to teach adolescents interview skills. Some studies have also looked at video-modeling, constant time delay, or in-situ training, in combination with BST, to help individuals gain safety skills and prevent potentially dangerous situations (Gast et al. 1993; Gunby, Carr, & LeBlanc, 2010; Gunby & Rapp, 2014; Miltenberger et al., 1999;).

Some studies have focused on teaching abuse and abduction prevention skills to individuals with autism or other developmental disabilities; many of these studies used behavioral skills training as their method of teaching (Fisher, Burke, & Griffin, 2013; Gunby, Carr, & Leblanc, 2010; Gunby & Rapp, 2014; Ledbetter-Cho et al., 2016; Lumley et al., 1998). There is not much literature that focuses on the use of other behavioral procedures to teach personal safety and even less which focuses on methods to teach these skills to individuals with developmental disabilities (Godish, Miltenberger, & Sanchez, 2017; Kurt & Kutlu, 2019). Kurt and Kutlu (2019) found the use of social stories to be effective in teaching children with autism abduction prevention skills. Godish, Miltenberger, and Sanchez (2017) found the use of video modeling to be effective in teaching children with autism how to respond to lures from strangers. Other approaches, such as the use of commercially available training programs, have been ineffective at teaching children the intended safety skill (Beck & Miltenberger, 2009).

While the procedures of these studies may differ, nearly all studies on preventing abuse or abduction measure their effectiveness using naturalistic or “in situ” assessments. For abuse prevention studies, naturalistic assessments typically took place in an environment where abuse could potentially occur (e.g. a group home). In these assessments, a confederate posing as a staff member made an inappropriate or sexual request to the study participants. The participants’ responses were scored according to how many components of a safety response they emitted. In abduction prevention studies, a naturalistic assessment is typically conducted by leaving the client alone somewhere (e.g. outside school, a grocery store, outside their home) where they are then approached by a confederate posing as a stranger. Confederates then deliver a lure for the child to come with them. After the lure, the child’s response is scored based on how many components of the safety response the child performed. In all naturalistic probes, the confederate either left the situation or another individual intervened if the participant agreed to the confederates’ lure or request.

According to Johnson et al. (2005), there are four lure types identified by researchers which are commonly used by abductors; simple, authority, and incentive were the three identified by Poche, Brouwer, and Swearinger (1981) and the assistance lure was identified by Holcombe, Wolery, and Katzenmeyer (1995). The simple lure is when the confederate requests a person to come with them for no particular reason (e.g. “Do you want to go for a walk with me?”). The authority lure occurs when the confederate tells the child that someone with authority (typically a parent or teacher) said it was okay to leave with him (e.g. “Your mom asked me to pick you up from school”). Incentive lures involve the confederate promising the child a reward for leaving with him or her (e.g. “If you come with me, you can play on my iPad”). Finally, in an assistance lure, the confederate asks the child for help with something (e.g. “I lost my puppy, can you help me find it?”). The use of behavioral skills training and in situ assessments are key components seen throughout behavior analytic personal safety literature.

Theoretical Underpinnings

For those practicing in the field of behavior analysis, certain criteria should be met when conducting behavior analytic research or treatment. These criteria are known as the seven dimensions of ABA and were first defined by Baer, Wolf, and Risley (1968). Baer, Wolf, and Risley (1968) stated that applied behavior analysis should be applied, behavioral, analytic, technological, conceptually systematic, effective, and should produce appropriately generalized outcomes (generality). These seven dimensions are what practicing behavior analysts should use to guide all treatment decisions. Each of the seven dimensions will be discussed below in relation to the topic of prevention of abuse and abduction.

Applied: The applied dimension states that applied behavior analysis should be practiced in the real world to solve real, socially significant problems. Baer, Wolf, and Risley (1968) state that the applied dimension focuses on problems that are important to people and society. Abuse and abduction are significant problems faced by our society today, as highlighted by the statistics presented in the introduction. Prevention of these events can increase an individual’s sense of personal safety, quality of life, and can even be lifesaving. Safety skills and prevention are a well-established applied area of behavior analysis. Many studies focusing on abuse or abduction prevention have also included questionnaires for caregivers or participants to indirectly assess the value and acceptability of interventions (Beck & Miltenberger, 2009; Kurt & Kutlu, 2019; Lumley et al., 1998).

Behavioral: The behavioral dimension involves focusing on behaviors that are observable and measurable. The behavioral dimension is composed of three main points. First, the behavior chosen for study must be the behavior in need of improvement. Second, the behavior must be measurable. Third, when changes in behavior are observed, it is necessary to ask whose behavior has changed.

Bandura (1977) asserted that skill-enhancing approaches that evaluate the actual performance of the skill are superior to those in which symbolic modeling is used. In research regarding abduction and abuse prevention, many of the articles directly measure the behavior in need of improvement through in situ assessment. Studies focusing on abduction prevention measured the child’s response to lures in rehearsal and in test situations designed to imitate an abduction attempt (Beck & Miltenberger, 2009; Lumley et al., 1998; Miltenberger et al., 2013). Research has shown that there is often a lack of correspondence between a participant’s knowledge and skills, making it important not to rely solely on knowledge measures (Carroll-Rowan & Miltenberger, 1994; Lumley & Miltenberger, 1997). The use of direct assessment of skills is what qualifies these studies as behavioral.

How the researchers measure the target response relates to the second point of the behavioral dimension; the behavior must be measurable. In most studies looking at the prevention of abuse or abduction, the target response is typically measured using a points system. The participant scores a range of points depending on whether they correctly emitted part or all of the target response. Haseltine and Miltenberger (1990) rated participants safety response on a 0- to 3- point scale: 0 = participant went with the abductor, 1 = participant did not go with abductor but did not move away or verbally refuse, 2 = participant either verbally refused, told a staff member, or moved 6.5 m away, and 3= the participant refused, left, and told a staff member.

The final point, “Whose behavior has changed?”, is evaluated by not only ensuring the last two points but by evaluating the observer or implementers’ behavior. This is often done in study and practice through thorough training of implementers, integrity checks, and measures of inter-observer agreement. Some studies not only took treatment integrity data for training sessions to ensure all training components were implemented, but also took treatment integrity data on the naturalistic assessments. Kurt and Kutlu (2019) took integrity data on these assessments by looking at whether (a) the participant was left alone, (b) the stranger delivering the abduction lure waited in a convenient position, (c) the lure was presented appropriately, (d) the participant was reinforced for correct responding by the second author, and (e) stranger delivering lure moved away when the participant responded incorrectly. Only after ensuring that procedures were implemented with fidelity can we presume that behavior change was a result of the intervention.

Analytic: In applied behavior analysis a study is analytic when a functional relationship has been demonstrated between the independent variable and the targeted behavior. In single study research, a study can be demonstrated as being analytic by the design used. One common design used in single-subject research is a reversal design. A reversal design demonstrates functional relation by showing consistent changes in the target behavior to the independent variable being added, changed, or removed. However, when looking at studies done on safety skills, it is usually not possible to use a reversal design. The reason for this is that once a skill is taught there is no way to “unteach” the skill. In cases such as these, a multiple baseline design can be used to demonstrate a functional relation. A multiple baseline design involves replicating the results of the intervention across multiple participants, settings, or behaviors. Much of the literature focusing on abduction and abuse prevention uses a multiple baseline across participants design to demonstrate a functional relation (Godish, Miltenberger, & Sanchez, 2017; Gunby, Carr, & LeBlanc, 2010; Vanselow & Hanley, 2014).

Technological: The technological dimension of applied behavior analysis refers to all procedures being defined in enough detail so that the reader could replicate the methodology and achieve the same results. Baer, Wolf, and Risley (1968) state that it is insufficient to state what should be done when the client emits the target response, but also what should be done if the client makes any alternative responses. In studies on abduction and abuse prevention, many methodologies not only list what to do when the client emits the target safety response, but what to do when they do not emit the safety response or only parts of it.

Conceptually Systematic: In behavior analytic research, procedures should be described in terms of the relevant basic principles from which they were derived, which is known as being conceptually systematic. The purpose of this is to allow those reading the research to derive other, similar procedures from the same principle and to become an integrated discipline that can be expanded, learned, and taught (Cooper, Heron, & Heward, 2007). As previously stated, much of the research on abuse and abduction prevention is taught using behavioral skills training. Erhard, Falcomata, and Harmon (2019) explain each of the components of behavioral skills training, instruction, modeling, rehearsal, and feedback, in a manner that can be tied to the basic principles from which they were derived.

Instruction involves telling the client each step of the behavior, the circumstances in which the behavior should be emitted, and the reinforcers which will be provided contingent on completing the target behavior (Erhard, Falcomata, & Harmon, 2019). Instructions can be tied to basic principles in that they serve to operationally define the target behavior, be a task analysis of the steps involved, define the relevant environmental variables that should be present when performing the target behavior, and define the contingency for performing the target behavior.

Modeling involves demonstrating the skill to the learner, which can be done live or symbolically (Cooper, Heron, & Heward, 2019). Symbolic models can include pictures, videos, audio, or a mixed media presentation of the desired behavior. For modeling to be effective, the client must already be able to imitate others. Studies have shown that symbolic modeling such as video modeling can be effective for teaching skills to children with autism and may potentially be more effective than live modeling (Godish, Miltenberger, & Sanchez, 2017; Kurt & Kutlu, 2019).

The third component of behavioral skills training, rehearsal, is when the client performs the target behavior so that instructors can confirm the acquisition of the target behavior, determine whether the behavior or components are incorrect, and/or provide feedback. While modeling helps demonstrate the behavior to participants, skill acquisition is achieved by performing the behavior (Bandura, 1977). Performing behavior in role-play helps to shape participants’ performance until criterion responding is achieved.

The final step, feedback, ties to the basic principles of reinforcement. Praise or other reinforcers are provided for the client when they correctly perform or attempt to perform the target behavior. Reinforcement in these studies is often provided in the form of verbal praise. The instructor also provides feedback on areas performed incorrectly. Clients often are provided corrective feedback in the form of additional instruction, modeling, and/or rehearsal until the client has demonstrated they can correctly perform the target behavior.

Effective: Another dimension of applied research is its effectiveness. The effectiveness of an intervention can be demonstrated by the extent to which is produces socially significant behavior change. Behavior analysis has produced significant behavior change in the area of abuse and abduction prevention. Other methods, such as commercially available training programs, have been demonstrated to be ineffective in creating behavior change when administered without additional training (Beck & Miltenberger, 2009; Miltenberger et al., 2013; Poche, Yoder, & Miltenberger, 1988; Wurtele et al., 1986). Many studies also indirectly looked at the social significance of the intervention by surveying participants, parents, or caregivers on whether they felt the intervention was effective and beneficial (Egemo-Helm et al., 2007; Ledbetter-Cho et al., 2016).

Generality: Generality in applied behavior analysis refers to the ability for behavior change to last over time, appear in a variety of environments, and/or spread to other behaviors not directly targeted by the intervention. Many applications in these areas have struggled with promoting generality despite training with multiple exemplars of lures and strangers (Collins, Schuster, & Nelson, 1992; Lumley et al., 1998). Research has found that in-situ training may be a necessary component to promote generalization of safety skills for some participants. In studies that have demonstrated generality across settings, participants were assessed on their ability to perform the same target response in a variety of untrained environments or locations (Beck & Miltenberger, 2009; Kurt & Kutlu, 2019; Poche, Yoder, & Miltenberger, 1988). Studies that have demonstrated generality over time typically conduct follow-up assessments several weeks to several months after the initial training to see whether the participant is still able to demonstrate the correct target behavior (Beck & Miltenberger, 2009; Poche, Yoder, & Miltenberger, 1988; Wurtele et al., 1986). Generality is extremely important in abuse and abduction prevention. It is unlikely that clients who are unable to generalize skills to untrained environments would demonstrate the safety response in a real abuse or abduction attempt.

As discussed, the seven dimensions of behavior analysis are present in much of the research on personal safety. Before Poche, Brouwer, and Swearinger’s 1981 article much research on this topic was not behavior analytic. By taking the seven dimensions and applying them to the study of personal safety, the field of behavior analysis has been able to make great strides in teaching these skills. While much has already been done to improve how we teach personal safety to children, the seven dimensions should also be applied moving forward with future research and applied work.

Applications

Since Poche, Brouwer, and Swearinger’s 1981 article, many advancements in personal safety research have been made. Many researchers have analyzed how best to address this societal problem of personal safety. Research has assessed a variety of interventions such as behavioral skills training, social stories, video modeling, and in situ training. Participants have also varied in terms of age, sex, and diagnoses. Research has also been done in a variety of settings such as group homes, schools, and various community settings. This section of the paper will evaluate, compare, and contrast applications in the area of personal safety for those with developmental disabilities. These applications will be reviewed in terms of the participants, methods, and results to summarize the findings of this body of research and bring up issues for future research.

Research on personal safety has achieved socially significant results using a variety of teaching methods. More recently in the literature, these methodologies have been examined in terms of their effectiveness in teaching personal safety skills to individuals with disabilities. Although teaching abduction prevention and abuse prevention skills may differ in methodology and response definitions, they are both evaluated in this paper since they both can be considered personal safety skills. Additionally, methods for assessing and teaching abuse-prevention skills, as discussed in Lumley and Miltenberger (1997), were largely based on the research done on abduction prevention for individuals with developmental disabilities.

Abduction Prevention

Behavioral skills training.

As previously mentioned, behavioral skills training is one of the most common methodologies for teaching abduction and abuse prevention. Multiple studies have found behavioral skills training to be effective in teaching abduction prevention skills to individuals with disabilities. Collins, Shuster, and Nelson (1992) found that behavioral skills training with the use of constant time delay procedures was effective in teaching adults with disabilities abduction prevention skills during role-play. Constant time delay involves immediately prompting the response after the instruction when first teaching a skill. Once the learner becomes more proficient in the skill, the time delay from the instruction to the response prompt is gradually increased. The authors stated that the purpose of constant time delay is to help minimize errors and promote rapid acquisition of skills. Despite trying to train for generalization through using multiple strangers, lures, settings, times, and trainers, these skills did not consistently generalize to in situ assessments.

Gast et al. (1993) expanded on the findings of Collins, Schuster, and Nelson (1992) by evaluating the effectiveness of behavioral skills training with constant time delay, with the addition of in situ training for teaching preschool children with disabilities abduction-prevention skills. Gast et al. (1993) used in situ training to help improve skill generalization, which was not demonstrated in Collins, Schuster, and Nelson (1992).  Results of Gast et al. (1993) found that that the safety response was quickly acquired in training but did not fully generalize to naturalistic probes. In situ training was then implemented to help promote generalization. Following in situ training three of four children met the criterion. The fourth participant continually exhibited the critical motor response but not the vocal protest. Those who participated in maintenance assessments all maintained the previous level of responding for two months in follow-up probes. These results replicated previous findings that behavioral skills training is effective in teaching safety skills but do not always promote generalization, even when trained with multiple exemplars. The use of constant time delay may have increased the efficacy of the training program. However, this study does not evaluate whether constant-time-delay procedures are more effective than other teaching or prompting methods.

Gast et al. (1993) demonstrated the effectiveness of adding in situ training after an intervention. In situ training is training done in the natural setting where the behavior is supposed to occur. During in situ training, the behavior is often practiced using instruction, modeling, and rehearsal this practice typically is done after the client emits an incorrect response on a naturalistic assessment. As demonstrated in Gast et al. (1993) in situ training may help to promote generalization and maintenance of skills. Several other studies have used in situ training as part of their intervention for this reason (Bergstrom, Najdowski, & Tarbox, 2014; Fisher, Burke, & Griffin 2013; Gast et al., 1993; Gunby, Carr, & LeBlanc, 2010; Gunby & Rapp, 2014).

Gunby, Carr, and LeBlanc (2010) evaluated the use of behavioral skills training with in situ training to teach abduction-prevention skills to three boys diagnosed with autism. Gunby, Carr, and LeBlanc (2010) found that behavioral skills training was effective in teaching the skills for one participant, but in situ training was necessary for skill acquisition for the other two participants. All participants maintained the skills at follow-up which occurred to three to seven weeks after posttraining. The results of Gunby, Carr, and LeBlanc (2010) raise the question of whether initial behavioral skills training was a necessary component for an effective intervention or if in situ training alone would have been equally effective in teaching all participants abduction prevention skills.

Fisher, Burke, and Griffin (2013) also evaluated the use of behavioral skills training and in situ training in teaching abduction prevention to young adults with disabilities. This article sought to address some of the methodological limitations of in situ research. One such limitation is that during in situ training, the trainer often makes repeated interruptions during naturalistic assessments. Fisher, Burke, and Griffin (2013) state that these interruptions can lead to the participant anticipating the arrival of the instructor, putting the participant at risk of failing to respond appropriately to an actual abduction lure. Another weakness this study addressed was that in situ assessments typically only occurred at baseline and posttraining rather than throughout training. Continual assessment of skills can demonstrate whether generalization is occurring during training. Fisher, Burke, and Griffin (2013) addressed both of these limitations in their study and found in situ training helped participants acquire the skills, but not all participants were able to maintain criterion responding. However, all participants met criterion responding for most follow-up sessions which occurred once per month for three months.

Gunby and Rapp (2014) evaluated the effects of behavioral skills training and in situ feedback on teaching abduction-prevention to three participants diagnosed with autism ages five and six. The purpose of this study was to further evaluate the effectiveness of behavioral skills training and in situ feedback on teaching abduction prevention skills to children with autism. However, this study also used a high-p request sequence before each lure. High-p request sequences involve asking the participant to complete several easy demands they have a history of complying with before presenting a more difficult demand. Authors state the purpose of the high-p request sequence was to simulate a grooming or recruitment process often used by abductors. Results of Gunby and Rapp (2014) found that BST alone did not produce criterion responses during posttraining, but all participants met criteria after in situ training.

Bergstrom, Najdowski, and Tarbox (2014) built off the findings of Gunby, Carr, and LeBlanc (2010) and Gunby and Rapp (2014) by replicating the use of behavioral skills training and in situ training to teach children with autism abduction-prevention skills. Bergstrom, Najdowski, and Tarbox (2014) addressed two weaknesses of these studies by using male abductors and focusing on stimulus generalization to untrained environments and abductors. The use of male abductors is an important component since 95% of nonfamily abductions are perpetrated by males (Finkelhor, Hammer, & Sedalk, 2002). Another thing that made this study different from previous research is that only the incentive lure was used. Indirect assessments were used to identify preferred items to use in the lure for each child. Bergstrom, Najdowski, and Tarbox (2014) found that all participants acquired the skills relatively quickly (3-7 sessions) during training. In situ training needed to be conducted for one participant during posttraining. However, during posttraining, this participant became very excited to see the therapist during in situ training. To avoid the therapist’s presence reinforcing incorrect responding, an incentive was added for correct responding for this participant. After the incentive, his performance returned to the criterion and the incentive reminder was faded.

Ledbetter-Cho et al. (2016) evaluated the effectiveness of behavioral skills training on improving the abduction-prevention skills of children with autism. Participants in this study were four males with autism, ages nine to eleven. The behavioral skills training included a rationale, instructions, video modeling, rehearsal, praise, and feedback. Behavioral skills training improved all four participants’ abduction-prevention skills which generalized to post-training assessments in new settings for three of four participants. Two participants maintained the skills at four weeks follow up. While most participants learned and demonstrated the skills during training sessions, few participants demonstrated the skills during generalization probes and follow-up.

Fisher (2014) demonstrated that individuals with Williams Syndrome could be taught abduction-prevention skills using behavioral skills training. Before intervention, one-third of all 21 participants agreed to go with the stranger and only three participants said no and walked away. After behavioral skills training participants were able to correctly perform 96% of steps during role play. In situ assessments found that intervention increased the use of safety skills, but they were not consistently demonstrated for all participants. This reiterates past findings which suggest little generalization between self-reported behavior or role-play and in situ probes.

Video modeling.

Godish, Miltenberger, and Sanchez (2017) took a different approach and evaluated the effectiveness of video modeling for teaching abduction prevention skills to children with autism. Although video modeling was used in Gunby, Carr, and LeBlanc (2010) it was part of a larger behavioral skills training treatment package. Godish, Miltenberger, and Sanchez (2010) aimed to look at the use of video modeling alone, with the use of in situ training as necessary. The results of this study found video modeling to be effective at teaching all four participants abduction prevention skills. Variable results were seen for one participant during follow-up. This variability persisted despite in situ training and added incentive. The authors state that this was likely due to the reactivity of the participant being aware he was being assessed.

Godish, Miltenberger, and Sanchez (2017) was one of the first abduction-prevention studies to find an antecedent-only approach to be effective at teaching safety skills. Previous studies such as Beck and Miltenberger (2009) and Miltenberger et al. (2013) found video programs intended to teach stranger safety skills ineffective. Godish, Miltenberger, and Sanchez (2017) state that one difference between this study and previous studies was that the participants were children with autism. Based on evidence from previous studies using video modeling it may be that video modeling is an effective intervention for children with autism. It may also be the video content that accounted for these different results. The authors state that the specificity of the video and robustness of the multiple exemplars may be what made video modeling an effective intervention in this case.

Social stories.

Kurt and Kutlu (2019) is another application that found an antecedent approach effective in teaching abduction prevention skills to children with autism, this time using social stories. Kurt and Kutlu (2019) read three- to five-minute-long social stories and asked comprehension questions prior to a naturalistic probe. Once the child emitted the correct response after a session, fading occurred to increase the length of time between reviewing the social story and the naturalistic probe. Additionally, these safety skills were maintained at one- and four- weeks follow-up. The use of social stories may be preferred to behavioral skills training due to the use of less intrusive prompts, less training time for practitioners, and less expertise needed for implementation.

Abuse Prevention

Behavioral approaches.

Following Lumley and Miltenbergers (1997) article, “Sexual Abuse Prevention for Persons with Mental Retardation”, which pointed out the lack of behavioral studies on sexual abuse prevention. Lumley et al. (1998) conducted the first empirical study on directly teaching and assessing sexual abuse prevention skills. In this study, six adult women diagnosed with mental retardation were taught sexual abuse prevention skills through behavioral skills training. Following behavioral skills training, these women were able to demonstrate the skills during knowledge assessments, verbal reports, and role-play, but failed to generalize the skills to the naturalistic probes. One year later a follow up to Lumley et al. (1998) was conducted by Miltenberger et al. (1999). Miltenberger et al. (1999) sought to enhance the generalization of these safety skills by following behavioral skills training with in situ training. The results of this study found that the behavioral skills training resulted in participants gaining the skills, while in situ training helped generalize the skills to naturalistic assessments.

Egemo-Helm et al. (2007) expanded on the research of Lumley et al. (1998) and Miltenberger et al. (1999) by evaluating behavioral skills training in combination with early in situ training to teach abuse prevention skills to women diagnosed with mental retardation. The results of this study found that, for most participants, the addition of early in situ training led to faster skill generalization relative to the results from Miltenberger et al. (1999). The use of early in situ training increased the efficiency and effectiveness of intervention for most participants, therefore decreasing the participants’ amount of exposure to simulated sexual abuse scenarios during assessment.

Kim (2016) found that the components of behavioral skills training, in combination with other teaching methods, were effective in teaching sexual abuse prevention skills to three elementary-aged girls with intellectual disabilities. While Kim (2016) did not directly refer to the training as behavioral skills training, the intervention was comprised of all the necessary components. Kim (2016) found the intervention to be effective at teaching the skills which also generalized to in situ assessments and was maintained for 10 weeks. However, due to the use of other teaching strategies (e.g. posters and story cards) we cannot be sure what components of the intervention were most effective for teaching the skills.

Cognitive approaches to physical abuse.

Very few applications can be found on prevention of physical abuse in the behavior analytic research. Most of the behavior analytic research available on abuse prevention focuses more on abduction and sexual abuse prevention. Doughty and Kane (2010) reviewed past studies on abuse prevention and pointed out the work being done in other fields on prevention of physical abuse such as the research of Dr. Khemka which utilizes a cognitive approach to abuse prevention. The work of Khemka and colleagues has focused on applying a cognitive decision-making approach to the areas of sexual, physical, and verbal abuse prevention skills.

Studies conducted by Ishita Khemka and colleagues demonstrate that a cognitive decision-making approach and related curriculum is effective in increasing participants independent decision making in regard to situations of abuse (Hickson et al. 2015; Khemka, 2000; Khemka & Hickson, 2000; Khemka, Hickson, & Reynolds, 2005). Outcomes of these trainings were measured by presenting short video vignettes demonstrating a situation of abuse and measuring participants verbal responses to a structured interview questionnaire on what the “decision maker” in the vignette should do or what they should do if they were in the scenario. Results of these studies demonstrate that a cognitive decision-making approach was effective in increasing appropriate, independent decision making. However, these studies only measure participants’ verbal reports of what they or the “decision maker” should do. Future studies should evaluate whether these skills can be demonstrated in role-play or in-situ assessments. Combining a cognitive intervention approach with behavioral assessment strategies could help further abuse prevention research.

Summary of Findings and Gaps in Current Research

Most of the major findings in the areas of abuse and abduction prevention for individuals with disabilities have been outlined above. However, there are certain weaknesses to this body of literature that must be addressed. Firstly, behavior analytic research on prevention of physical abuse is scarce. Additionally, the number of studies using interventions other than behavioral skills training are limited. The more recent studies done by Godish, Miltenberger, and Sanchez (2017) and Kurt and Kutlu (2019) have found other interventions to be effective in teaching safety skills. While behavioral skills training is an effective treatment when combined with in situ training, it can be time-consuming and difficult to apply to many settings. Programs utilizing social stories and video modeling may be more efficient and easier for parents and other caregivers to implement.

While other methods of teaching personal safety skills are less common in the literature, this may be due to publication bias. Most all the articles reviewed resulted in the client gaining some level of knowledge or skills. It is possible that many other studies were conducted but not published due to the intervention being ineffective. Publication bias is a factor that should be considered when judging the efficacy of an intervention. Although publication bias could be at play, variations in interventions and methodology should still be explored in future research.

One suggestion that repeatedly appeared in many articles was the need to conduct a component analysis of intervention packages. Behavioral skills training contains multiple intervention components, some of which may not be necessary to achieve effective results. Previous studies utilizing behavioral skills training to teach abduction prevention found that giving a rationale is not a necessary component for effective intervention and may lead to negative side effects, such as fear, for participants (Poche, Brouwer, & Swearinger, 1981). Other elements of these treatment packages may also not be necessary. For example, Miltenberger et al. (2013) found that in situ training alone was effective in teaching abduction-prevention skills to children. By figuring out which components are necessary to produce significant results, we can come up with treatments that are efficient and effective. In addition to analyzing which components of behavioral skills training are effective, future research should focus on analyzing the long-term maintenance of these interventions and increasing this maintenance by providing booster sessions.

Another way that future research can be improved is by using approaches such as Gunby and Rapp’s (2014) in which a high probability request sequence was presented before the lure. Future studies could focus on ways to simulate the participants having a relationship with the abuser or abductor. Poche, Yoder, and Miltenberger (1988) pointed out the need to take the emphasis off the word “stranger” when teaching abduction prevention skills. This is important since many child abductors are somewhat known to their victims. By using high-probability request sequences or pairing the confederate with the participants’ daily activities, we may get a better indication of how the child would respond to a real abuse or abduction scenario. Future research should focus on a way to ethically simulate previously established relationships the participants may have with potential abusers or abductors.

Participant demographics are another area in which gaps can be seen within the literature. Findings of these studies need to be extended to participants of certain sexes, age groups, and ability levels. Studies focusing on abuse prevention have all recruited individuals diagnosed with mental retardation or intellectual disabilities. In all four studies, participants were stated to have mild to moderate intellectual disabilities. Most of these studies, 3 of 4, recruited adult participants while the fourth study focused on children ages 11-13. All abuse prevention studies recruited female participants.  Future research in this area should focus on expanding research by recruiting male participants, teens, and younger children. While studies show that women are more likely to be victims of abuse than men, men and boys can still be victims of abuse. While current research on rates of abuse among males with disabilities is scarce, studies have shown an overall increased likelihood of abuse for persons with disabilities (Hughes et al., 2012; Jones et al., 2012). Additionally, a brief published by Vera Institute of Justice reported that men with disabilities and deaf men were twice as likely to experience violent victimization such as sexual assault, robbery, and aggravated assault (Hastings & Harrell, 2017).

Of the applications teaching abduction-prevention skills: six studies focused specifically on children with autism, one focused on children with disabilities characterized as having “developmental delays”, one study focused on adults with intellectual disabilities, and one study on adults with Williamson Syndrome (a developmental disability characterized by intellectual disability as well as increased sociability towards strangers). Ability level was difficult to gauge from these studies, since many different measures and criteria were used to assess this. Many participants in these studies either did not have significant intellectual or language impairments or were classified as having mild-moderate intellectual disabilities. Some studies, such as Godish, Miltenberger, and Sanchez (2017), directly state the ability level of participants. Other studies indicate this through requirements for participation. For example, Bergstrom, Najdowski, and Tarbox (2014) required that participants be able to exhibit an extensive vocal-verbal repertoire, knowledge of the definition of a ‘stranger’, ability to follow instructions and rules, and ability to learn through role-play.

Abduction prevention studies have focused on a variety of ages. However, a gap existed for children ages 12-16; which is a group that may be more susceptible to abduction. According to NCMEC’s “Analysis of Attempted Abduction Trends” data from 2005 to 2019, child abduction was most common with children between the ages of 10 to 14 (The National Center for Missing and Exploited Children, n.d.). Additionally, the NCMEC’s data found that abductions of older children were more likely to include a sexual component such as indecent exposure or sexual assault (The National Center for Missing and Exploited Children, 2016). Future studies should focus on applying interventions to this adolescent age range or evaluate how interventions at younger ages maintain over the years.

While both sexes were represented in the abduction-prevention literature, sexes in these studies tended to reflect diagnosis demographics. Studies focusing on individuals with autism recruited all or mostly male participants. Recent studies have found that there likely exist marked differences between males and females diagnosed with autism spectrum disorder. Due to these differences, questions have been raised as to whether intervention technologies studied with male participants can or should be applied to female participants. Future studies should focus on seeking out female participants for inclusion in these studies (Shefcyk, 2015).

Although teaching personal safety skills is an area that applied behavior analysis has worked to address for several years, gaps in the literature still exist. The current gaps in the literature include few behavioral studies on physical abuse prevention, few studies utilizing interventions other than behavioral skills training, the need for component analyses for package interventions, more applications focusing on familial abductions/known abductors, as well as gaps related to participant demographics (particularly age, sex, and ability). Future research should work on addressing the previously mentioned gaps in the literature.

Ethical Considerations

The previous section looked at teaching personal safety skills to individuals with disabilities in the context of applied research. These applications went through institutional review boards for approval before being implemented to ensure the studies were ethically sound and that precautionary measures were in place. Prevention of abuse and abduction, while important, can be a sensitive skill area to teach in the applied setting. While there is no formal review process for everyday applied practice, ethical concerns will likely arise when teaching skills relating to sensitive topics. Practitioners have ethical obligations in relation to teaching these skills in a manner that benefits the client and reduces possible risks of harm. These ethical obligations will be discussed in relation to the Professional and Ethical Compliance Code for Behavior Analyst’s (PECC), cost-benefit analyses, and ethical decision making.

Behavior Analysts must follow the ethics code set forth by the Behavior Analyst Certification Board (BACB) which is the presiding code for ethical decision making the field of behavior analysis (Behavior Analyst Certification Board, 2012). The code consists of many items designed to promote ethical practice in the field, many of these items can be directly related to teaching personal safety skills in the applied setting. However, when reviewing the code in relation to teaching personal safety skills, many items seem to contradict each other. Ethical decision making, based on the PECC, should be used to determine whether teaching personal safety skills is necessary for that client and what is the best way to approach intervention. The following PECC codes should be considered when teaching personal safety skills: 1.04 Integrity, 2.09 Treatment/Intervention Efficacy, 2.15 Interrupting or Discontinuing Services, 4.02 Involving Clients with Planning and Consent, and 4.05 Describing Behavior Change Programs and Objectives.

Code 1.04, Integrity, is an important code to consider when teaching personal safety skills. This code has several subsections describing how behavior analysts maintain integrity in practice. Section A of code 1.04 states, “behavior analysts are truthful and honest and arrange the environment to promote truthful and honest behavior in others” (Behavior Analyst Certification Board, 2012). This code can come in contradiction with teaching personal safety skills due to the use of deceit during assessment. Practitioners may choose to avoid the use of deceit altogether by only assessing these skills during role-play or verbal reports. However, these methods do not always accurately reflect how a client would react in response to a real abuse or abduction lure, indicating that deceit may be necessary for assessing the effectiveness of intervention.

Although the use of deceit seems to contradict code 1.04. Code 2.09 (a) states that clients have the right to effective treatment (Behavior Analyst Certification Board, 2012). Treatment should be based on effective applications and individualized to that client. In applications on teaching personal safety skills, most all the studies assessed treatment efficacy using naturalistic assessment. While codes 1.04 and 2.09 seem to contradict each other in this case, it is important to note that in such an assessment, deceit is not being used for any malicious purposes. Finding ways to mitigate the use of deceit, such as gaining consent before implementation and debriefing the client after each assessment, are some options for ways to maintain integrity while analyzing the effectiveness of the intervention.

Another portion of the ethics code to consider when planning to teach abuse and abduction prevention is 2.15 Interrupting or Discontinuing Services (Behavior Analyst Certification Board, 2012). Part D of this code states that behavior analytic services should be discontinued when the client no longer needs the service, is not benefitting from the service, is being harmed by continued service, or when the client requests discontinuation (Behavior Analyst Certification Board, 2012). This code is important in relation to safety skills due to potential negative outcomes that could occur, such as developing a fear of strangers. Several applications used side effects questionnaires to assess any negative outcomes of intervention and discontinued services in cases where negative side effects were seen (Godish, Miltenberger, & Sanchez, 2017; Kim, 2016). Due to the sensitive nature of abduction and abuse it is important for clinicians to develop a way to monitor for negative side effects and discontinue or modify intervention should these side effects develop.

Code 4.02, Involving clients in planning and consent, is a necessary step with all intervention programs but is especially important in the creation of interventions addressing sensitive topics (Behavior Analyst Certification Board, 2012). With topics such as abuse and abduction it is possible that clients have endured past traumas that need to be considered when planning an intervention. Also, as discussed previously, there is the potential for the client to develop new fears from the intervention, so gaining parent consent and client assent before each session is an especially important aspect of such interventions.

Along with gaining consent, it is important to describe intervention and assessment objectives, which are addressed by code 4.05 of the PECC (Behavior Analyst Certification Board, 2012). This code states that program objectives should be described in writing for the client before beginning to implement. Another important part this code mentions is the need to conduct a risk-benefit analysis of the procedures to be implemented. A risk-benefit analysis looks to see whether the benefits of intervention would outweigh possible risks. This is an important step to take before implementing abduction and abuse prevention skills due to the potential side effects that could occur.

None of the applications discussed conducted a formal cost-benefit analysis. However, the following are some factors to consider when determining whether to teach personal safety skills to a client: abuse and abduction risk factors, potential negative side-effects, and potential positive outcomes. One of the first considerations before conducting a cost-benefit analysis should be caregiver or client interest in a program that teaches personal safety skills. After the caregiver or client indicate this as an area of concern, it is then the job of the behavior analyst to determine whether learning personal safety skills is appropriate for the client, which skills to teach, and how to teach them.

Client demographics are one way to determine whether the client would benefit, more than most, from safety skills knowledge. People with disabilities are more likely to be victims of abuse than individuals without disabilities; however, other demographics may also increase their risk of abuse or abduction. Factors such as age, sex, living arrangements (e.g. group home or with parents), and level of independence are all factors that should be assessed in a risk-benefit analysis. A list of risk and protective factors for various forms of abuse can be found on the Center for Disease Control’s website under violence prevention (Center for Disease Control, 2020).

After determining risk factors for that client, potential benefits should be compared to potential risks of intervention. Positive intervention outcomes and prevention of future abuse and neglect should be compared against the probability of developing negative emotional responding. After conducting a risk-benefit analysis and considering the mentioned variables, behavior analysts should then determine which personal safety skills this client would benefit from and how they should be taught. It may be beneficial for some clients to receive general sex education, while for others, a behavioral skills training program for prevention skills may be needed. It is up to the behavior analyst to analyze potential risks and benefits and decide what is best for that client.

Rosenberg and Schwartz (2018) proposed an ethical decision-making process to help behavior analysts decide the most ethical route when addressing grey areas in the code or contradictions between multiple codes. As previously mentioned, when planning an intervention to teach abuse and abduction prevention multiple codes should be taken into consideration, some which seemingly contradict one another. According to Rosenberg and Schwartz (2018) ethical conduct is a skill that behavior analysts must regularly practice. The ethical decision-making process is one way to guide this practice. This model contains a series of steps to help behavior analysts identify and resolve ethical dilemmas.

There are 6 steps to ethical decision-making which are: (1) identify why this triggers your ethical radar, (2) brainstorm solutions, (3) evaluate possible solutions, (4) identify an acceptable solution or brainstorm more solutions, (5) implement the solution with fidelity and document all actions, and (6) reflect upon the results of your decision. Previously, it was explained how codes 1.04 (a) ‘integrity’ and 2.09 (a) ‘right to effective treatment’ may contradict each other when assessing abuse and abduction prevention skills. Following the ethical decision-making model for dilemmas like this and other possible ethical concerns can help a practitioner arrive at an ethical solution which benefits the client.

Many PECC codes apply when designing abuse and abduction prevention interventions, and it is important that practicing behavior analysts thoroughly know this code and apply it to all behavior analytic practice. Integrity, intervention efficacy, considerations for discontinuing services, client consent, and describing behavior change objectives are critical codes areas when teaching personal safety skills due to the sensitive nature of the topic and the potential for negative side effects to develop. Behavior analysts who plan to teach these skills to their client(s) should take these codes into consideration, conduct a risk-benefit analysis, and practice ethical decision making when designing and implementing intervention procedures.

Future Directions

Personal safety has long been an area of focus for behavior analytic research. Despite many years of study, there still exist limitations in the literature which should be addressed. Conducting component analyses of behavioral skills training packages, evaluating other interventions to promote personal safety, addressing familial abuse or abduction, and teaching skills to various age groups and sexes are all areas in which future research could expand this body of literature.

Not only should these gaps in the literature be investigated, but the topic of personal safety should regularly be revisited by researchers, since trends in abuse and abductions statistics change over the years. In the NCMEC’s “A 10-Year Analysis of Attempted Abductions and Related Incidents” it was reported that force was the top method used against children in child abductions (NCMEC, 2016). However, past abduction prevention literature cites the use of lures as the common method used against children, and is the method which was used in teaching abduction prevention skills throughout the literature (Bergstrom, Najdowski, & Tarbox, 2014; Fisher, Burke, & Griffin, 2013; Gunby, Carr, & LeBlanc, 2010; Gunby & Rapp, 2014). Future studies focusing on abduction prevention should take more recent findings into account and modify interventions accordingly. For example, since studies show that force is often used to abduct children, it may be more beneficial to have the target response to an abduction attempt to simply be to run away quickly or to yell loudly.

Two large gaps in the literature are the lack of applications focusing on physical abuse prevention and the lack of applications using participants with significant intellectual or communication impairments. The lack of applications in these areas may be caused by the difficulties faced in designing effective interventions. This section of the paper will discuss why these gaps in the literature may exist and potential ways to address these problems.

Doughty and Kane (2010) reviewed relevant empirical studies (behavioral and cognitive) on abuse-prevention for people with intellectual disabilities from 1997 to 2010. Their search only found six empirical studies that took place during this time frame (Lumley et al., 1998; Miltenberger et al., 1999; Egemo-Helm et al., 2007; Khemka, 2000; & Khemka et al., 2005). Of these studies, only Khemka (2000) and Khemka et al. (2005) addressed physical abuse in their study utilizing a cognitive decision-making training approach.  Doughty and Kane (2010) noted that no studies used a behavior analytic approach to address physical abuse prevention of individuals with intellectual disabilities. Despite Doughty & Kane (2010) addressing this gap in research, still in 2020 it is difficult to find empirical research on physical abuse prevention for individuals with developmental disabilities. A search of several databases and journals relating to behavior analysis or developmental disabilities did not result in any empirical studies on the topic. Future research should address this apparent gap in the literature. If empirical studies have been done on this subject, better dissemination of this information is needed as evident by the difficulty in finding these studies.

Another gap in the literature is the lack of empirical studies on teaching personal safety skills to participants with significant intellectual or communication impairments. Doughty & Kane (2010) noted this as a gap in the abuse-prevention literature, however this is also true of studies addressing abduction prevention. Individuals in the above applications typically fall in the mild to moderate range of intellectual impairments and many studies note the participants as having relatively good communication skills. Doughty & Kane (2010) noted that these studies may recruit participants with efficient communication skills, since the assessment and teaching procedures are verbally mediated and that teaching these skills through verbal means may be easiest. However, individuals with more severe impairments likely face greater risks to personal safety due to difficulties in recognizing and communicating potentially dangerous situations. Future research should work towards designing simplified interventions that can be delivered through non-verbal means. Due to the communication difficulties many individuals with developmental disabilities face, alongside with more recent findings that most abductors use physical force against their victims, it may be beneficial to focus on the non-verbal components of a safety-response such as leaving the potentially dangerous situation (NCMEC, 2016).

The need for behavioral analytic approaches to physical abuse prevention and the need for appropriate techniques to teach personal safety skills to individuals with severe impairments were limitations pointed out by Doughty and Kane (2010). Despite Doughty and Kane (2010) pointing out these limitations, little headway has been made in these areas over the past 10 years. This leads to one of two conclusion(s): either additional attention needs to be brought to these gaps in the research, or researchers in these areas are confronted with significant barriers to addressing these gaps. Future research on the topic of personal safety should focus on filling in the gaps to the literature previously mentioned and beginning to identify the significant barriers that will need to be overcome in order to do so. To tackle such a large societal issue, collaboration with other fields, such as those using a cognitive approach, may be necessary. In addition to tacking the problem on an individual level, behavior analysis will likely need to take a social-ecological approach to prevention and intervene on the relationship, community, and societal level as well.

Conclusion

Much progress has been made in the behavior analytic literature on teaching personal safety skills to individuals with disabilities. However, despite finding socially significant interventions to teach such skills, barriers remain. The generality of this body of research can be improved by addressing the previously mentioned gaps in the literature regarding age, sex, and ability. Methods to teach personal safety skills to individuals with severe intellectual and/or communication impairments still need to be further researched. Breaking down behavioral skills training using component analyses, evaluating other teaching methods, and working towards finding non-verbal ways of teaching these skills are all ways in which these applications may potentially be adapted to benefit learners. Additionally, ethical ways to teach and assess skills related to physical abuse prevention still must be explored. Outcomes demonstrated by Khemka and colleagues should be evaluated behaviorally through using role-play or in-situ assessments.

Victimization of individuals with disabilities continues to be a problem in our society, and research should continue to evaluate how to prevent this from occurring. Lack of progress in recent years suggests this problem may need to be approached from a different lens. The barriers encountered within this research need to be identified. Identifying these barriers, collaborating with experts in other fields, and taking a holistic social-ecological approach to personal safety are all ways in which our field may come closer to ensuring the personal safety of the at-risk individuals we serve.

References

Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1(1), 91-97.

Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice Hall

Beck, K. V., & Miltenberger, R. G. (2009). Evaluation of a commercially available program and in situ training by parents to teach abduction‐prevention skills to children. Journal of Applied Behavior Analysis, 42(4), 761-772.

Behavior Analyst Certification Board. (2014). Professional and ethical compliance code for behavior analysts. Littleton, CO: Author.

Bergstrom, R., Najdowski, A. C., & Tarbox, J. (2014). A systematic replication of teaching children with autism to respond appropriately to lures from strangers. Journal of Applied Behavior Analysis, 47(4), 861-865.

Braukmann, C. J., Fixsen, D. L., Phillips, E. L., Wolf, M. M., & Maloney, D. M. (1974). An analysis of a selection interview training package for predelinquents at achievement place. Correctional Psychologist, 1(1), 30-42.

Carroll-Rowan, L. A., & Miltenberger, R. G. (1994). A comparison of procedures for teaching abduction prevention to preschoolers. Education and Treatment of Children, 113-128.

Center for Disease Control. (2020). Child abuse & neglect: Risk and protective factors. Retrieved from https://www.cdc.gov/violenceprevention/childabuseandneglect/riskprotectivefactors.html

Collins, B. C., Schuster, J. W., & Nelson, C. M. (1992). Teaching a generalized response to the lures of strangers to adults with severe handicaps. Exceptionality, 3, 67–80.

Cooper, J., Heron, T. and Heward, W. (2007). Applied behavior analysis. 2nd ed. Upper Saddle River, Nj: Pearson.

Cooper, J., Heron, T. and Heward, W. (2019). Applied behavior analysis. 3rd ed. Pearson.

Davis, L. A. (2011). Abuse of Children with Intellectual Disabilities. The Arc.

Doughty, A. H., & Kane, L. M. (2010). Teaching abuse-protection skills to people with intellectual, disabilities: A review of the literature. Research in Developmental Disabilities, 31(2), 331-337.

Egemo‐Helm, K. R., Miltenberger, R. G., Knudson, P., Finstrom, N., Jostad, C., & Johnson, B. (2007). An evaluation of in situ training to teach sexual abuse prevention skills to women with mental retardation. Behavioral Interventions: Theory & Practice in Residential & Community‐Based Clinical Programs, 22(2), 99-119.

Erhard, P., Falcomata, T. S., & Harmon, T. (2019). Encyclopedia of autism spectrum disorders. In Encyclopedia of Autism Spectrum Disorders. New York, NY: Springer. Retrieved from https://link.springer.com/referenceworkentry/10.1007/978-1-4614-6435-8_102320-1.

Facts about Children and Violence. (2017, April 11). Retrieved from https://www.justice.gov/archives/defendingchildhood/facts-about-children-and-violence

Finkelhor, D., Hammer, H., & Sedlack, A. J. (2008). Sexually Assaulted Children: National Estimates and Characteristics. Juvenile Justice Bulletin.

Fisher, M. H. (2014). Evaluation of a stranger safety training programme for adults with Williams syndrome. Journal of Intellectual Disability Research, 58(10), 903-914.

Fisher, M. H., Burke, M. M., & Griffin, M. M. (2013). Teaching young adults with disabilities to respond appropriately to lures from strangers. Journal of Applied Behavior Analysis, 46(2), 528-533.

Forgione, A. G. (1976). The use of mannequins in the behavioral assessment of child molesters: Two case reports. Behavior Therapy, 7(5), 678-685.

Gast, D. L., Collins, B. C., Wolery, M., & Jones, R. (1993). Teaching preschool children with disabilities to respond to the lures of strangers. Exceptional Children, 59(4), 301-311.

Godish, D., Miltenberger, R., Sanchez, S. (2017). Evaluation of video modeling for teaching abduction prevention skills to children with autism spectrum disorder. Advances in Neurodevelopmental Disorders, 1, 168–175.

Groth, A. N., & Birnbaum, H. J. (1978). Adult sexual orientation and attraction to underage persons. Archives of Sexual Behavior, 7(3), 175-181.

Groth, A. N., & Burgess, A. W. (1980). Male rape: Offenders and victims. The American Journal of Psychiatry.

Gunby, K. V., Carr, J. E., & Leblanc, L. A. (2010). Teaching abduction‐prevention skills to children with autism. Journal of Applied Behavior Analysis, 43(1), 107-112.

Gunby, K. V., & Rapp, J. T. (2014). The use of behavioral skills training and in situ feedback to protect children with autism from abduction lures. Journal of Applied Behavior Analysis, 47(4), 856-860.

Haseltine, B., & Miltenberger, R. G. (1990). Teaching self-protection skills to persons with mental retardation. American Journal of Mental Retardation: AJMR, 95(2), 188-197.

Hastings, A., & Harrell, S. (2017). Sexual victimization of men with disabilities and deaf men: a national snapshot.

Hickson, L., Khemka, I., Golden, H., & Chatzistyli, A. (2015). Randomized controlled trial to evaluate an abuse prevention curriculum for women and men with intellectual and developmental disabilities. American Journal on Intellectual and Developmental Disabilities, 120(6), 490-503.

Holcombe, A., Wolery, M., & Katzenmeyer, J. (1995). Teaching preschoolers to avoid abduction by strangers: Evaluation of maintenance strategies. Journal of Child and Family Studies, 4(2), 177-192.

Hughes, K., Bellis, M. A., Jones, L., Wood, S., Bates, G., Eckley, L., … & Officer, A. (2012). Prevalence and risk of violence against adults with disabilities: a systematic review and meta-analysis of observational studies. The Lancet, 379(9826), 1621-1629.

Johnson, B. M., Miltenberger, R. G., Egemo‐Helm, K., Jostad, C. M., Flessner, C., & Gatheridge, B. (2005). Evaluation of behavioral skills training for teaching abduction‐prevention skills to young children. Journal of Applied Behavior Analysis, 38(1), 67-78.

Jones, L., Bellis, M. A., Wood, S., Hughes, K., McCoy, E., Eckley, L., … & Officer, A. (2012). Prevalence and risk of violence against children with disabilities: A systematic review and meta-analysis of observational studies. The Lancet, 380(9845), 899-907.

Khemka, I. (2000). Increasing independent decision-making skills of women with mental retardation in simulated interpersonal situations of abuse. American Journal on Mental Retardation, 105(5), 387-401.

Khemka, I., & Hickson, L. (2000). Decision-making by adults with mental retardation in simulated situations of abuse. Mental Retardation, 38(1), 15-26.

Khemka, I., Hickson, L., & Reynolds, G. (2005). Evaluation of a decision-making curriculum designed to empower women with mental retardation to resist abuse. American Journal on Mental Retardation, 110(3), 193-204.

Kim, Y. R. (2016). Evaluation of a sexual abuse prevention program for children with intellectual disabilities. Behavioral Interventions, 31(2), 195-209.

Kurt, O., & Kutlu, M. (2019). Effectiveness of Social Stories in Teaching Abduction-Prevention Skills to Children with Autism. Journal of Autism and Developmental Disorders, 49(9), 3807-3818.

Lang, R. A., & Frenzel, R. R. (1988). How sex offenders lure children. Annals of Sex Research, 1(2), 303-317.

Ledbetter-Cho, K., Lang, R., Davenport, K., Moore, M., Lee, A., O’Reilly, M., … & Falcomata, T. (2016). Behavioral skills training to improve the abduction-prevention skills of children with autism. Behavior Analysis in Practice, 9(3), 266-270.

Lumley, V. A., & Miltenberger, R. G. (1997). Sexual abuse prevention for persons with mental retardation. American Journal of Mental Retardation: AJMR, 101(5), 459-472.

Lumley, V. A., Miltenberger, R. G., Long, E. S., Rapp, J. T., & Roberts, J. A. (1998). Evaluation of a sexual abuse prevention program for adults with mental retardation. Journal of Applied Behavior Analysis, 31(1), 91-101.

Miltenberger, R. G. (2008). Behavior modification: principles and procedures (4th ed.). Belmont, CA: Michele Sordi.

Miltenberger, R. G., Fogel, V. A., Beck, K. V., Koehler, S., Shayne, R., Noah, J., … & Godish, D. (2013). Efficacy of the stranger safety abduction‐prevention program and parent‐conducted in situ training. Journal of Applied Behavior Analysis, 46(4), 817-820.

Miltenberger, R. G., Roberts, J. A., Ellingson, S., Galensky, T., Rapp, J. T., Long, E. S., & Lumley, V. A. (1999). Training and generalization of sexual abuse prevention skills for women with mental retardation. Journal of Applied Behavior Analysis, 32(3), 385-388.

Miltenberger, R. G., & Thiesse‐Duffy, E. (1988). Evaluation of home‐based programs for teaching personal safety skills to children. Journal of Applied Behavior Analysis, 21(1), 81-87.

Missing children on the Autism Spectrum. (2016). Retrieved February 23, 2020, from http://www.missingkids.com/content/dam/missingkids/pdfs/ncmec-analysis/2017ncmecautismspectrum.pdf

The National Center for Missing & Exploited Children. (n.d.). Analysis of Attempted Abduction Trends. Retrieved from file:///C:/Users/trs05/Downloads/Attempted Abduction trends with talking points_2019.pdf

The National Center for Missing & Exploited Children. (2016). A 10-Year Analysis of Attempted Abductions and Related Incidents. Retrieved from file:///C:/Users/trs05/Downloads/attemptedabductions10yearanalysisjune2016 (1).pdf

Panton, J. H. (1978). Personality differences appearing between rapists of adults, rapists of children and non-violent sexual molesters of female children. Research Communications in Psychology, Psychiatry & Behavior.

Poche, C., Brouwer, R., & Swearingen, M. (1981). Teaching self‐protection to young children. Journal of Applied Behavior Analysis, 14(2), 169-175.

Poche, C., Yoder, P., & Miltenberger, R. (1988). Teaching self‐protection to children using television techniques. Journal of Applied Behavior Analysis, 21(3), 253-261.

Rosenberg, N. E., & Schwartz, I. S. (2018). Guidance or compliance: What makes an ethical behavior analyst?. Behavior Analysis in Practice, 12(2), 473-482.

Shefcyk, A. (2015) Count us in: Addressing gender disparities in autism research.

Sobsey, R., & Varnhagen, C. (1988). Sexual abuse and exploitation of people with disabilities. University of Alberta, Developmental Disabilities Centre.

Sullivan, P.M. (2002). Violence against children with disabilities: Prevention, public policy and research implications. Paper commissioned for the National Conference on Preventing and Intervening in Violence Against Children and Adults with Disabilities.

Suzy Lamplugh Trust. (n.d.). What is personal safety?. Retrieved from https://www.ukconstructionmedia.co.uk/case-study/slt-personal-safety/

Vanselow, N. R., & Hanley, G. P. (2014). An evaluation of computerized behavioral skills training to teach safety skills to young children. Journal of Applied Behavior Analysis, 47(1), 51-69.

Watson, J. D. (1984). Talking about the best kept secret: Sexual abuse and children with disabilities. Exceptional Parent, 14(6), 15-16.

Wurtele, S. K., Saslawsky, D. A., Miller, C. L., Marrs, S. R., & Britcher, J. C. (1986). Teaching personal safety skills for potential prevention of sexual abuse: A comparison of treatments. Journal of Consulting and Clinical Psychology, 54(5), 688.