EFTA00157004.pdf

q Routledge Journal of Child Sexual Abuse ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.comiloirwcsa20 Validation of the Sexual Grooming Model of Child Sexual Abusers Georgia M. Winters , Elizabeth L. Jeglic & Leah E. Kaylor To cite this article: Georgia M. Winters , Elizabeth L. Jeglic & Leah E. Kaylor (2020): Validation of the Sexual Grooming Model of Child Sexual Abusers, Journal of Child Sexual Abuse, DOI: 10.1080/10538712.2020.1801935 To link to this article: httos://doLorg/10.1080/10538712.2020.1801935 IN Published online: 02 Oct 2020. 7 Submit your article to this journal Gr I the Article views: 33 View related articles I? View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.comiaction/journalInformation?journalCode=wcsa20 3502-033 Page I of 22 EFIA_00001663 EFTA00157004 JOURNAL OF CHILD SEXUAL ABUSE Mips://dolorg/10.1080/10518712.2020.1801935 14 Routledge Taylor & (randS Group Crad,...updatti Validation of the Sexual Grooming Model of Child Sexual Abusers Georgia M. Winter?, Elizabeth L Jeglie', and Leah E. Kaylorb 'School of Psychology, Fairleigh Dickinson University, Teaneck, NJ, USA; °Psychology Department, John Jay College of Criminal Justice, New York, NY, USA ABSTRACT ARTICLE HISTORY Sexual grooming has been deemed an integral part of the child Received 10 January 2020 sexual abuse process. However, there has yet to be a universally Revised 25 March 2020 accepted model for this process and, as a consequence, there is Accepted 19 May 2020 no clear understanding of which behaviors constitute sexual KEYWORDS grooming. One proposed model of in•person sexual grooming Sexual grooming; child outlined five stages of the process: 1) victim selection, 2) gaining sexual abuser; child sexual access and isolating a child, 3) trust development. 4) desensiti- abuse; sex offenses zation to sexual content and physical contact, and 5) mainte- nance following the abuse. The present study sought to validate this Sexual Grooming Model (SGM) and identify behaviors that may be employed during each stage of the process. First, a thorough review of the literature was conducted to generate a comprehensive list of sexual grooming behaviors (ri = 77). Second, 18 experts in the field completed a survey which asked them to rate the extent to which each of the five stages and potential grooming behaviors were relevant to the sexual grooming process. Results provided support for the SGM and produced 42 behaviors that were considered to be grooming tactics within these stages. From this, the first validated, com- prehensive model of in•person sexual grooming is proposed. The ankle concludes with a discussion of the implications and future directions in the field. Child sexual abuse (CSA) is a serious public health issue with an estimated lifetime prevalence ranging between 12-27% for girls and 4-5% for boys in the United States and Canada (Briere & Eliott, 2003; Canadian Centre for Justice Statistics, 2017; Finkelhor et al., 2015; Letourneau et al., 2018). In the United States, individuals incarcerated for sexual offenses comprise 12% of state inmate populations (Department of Justice, 2014). Notably, however, preva- lence rates published by criminal justice agencies often underestimate the severity of this problem due to low rates of victim disclosure and formal reporting of cases (Leclerc & Wortley, 2015; Sethi et al., 2013). While there are numerous reasons CSA may go undetected or unreported, it has been suggested that a perpetrators' manipulation of the victims before and after the CONTACT Georgia M. Winters georgiawinters82@gmailcom 0 School of Psychology, faldeigh Colddnson University, Teanedc NJ 07666 0 2020 bier & Francs 3502-033 Page 2 of 22 EFTA_00001664 EFTA00157005 2® G. M. WINTERS ET AL. abuse, known as "sexual grooming," may decrease the likelihood of its detec lion and disclosure (Van Dam, 2001). It is estimated that almost half of the cases of CSA involve some element of sexual grooming (Canter et al., 1998). While there has yet to be a universally agreed upon definition in the literature, the term sexual grooming typically refers to the process by which an offender skillfully manipulates a potential victim into situations in which sexual abuse can be more readily committed, while simultaneously preventing disclosure (Van Dam, 2001; Wyre, 2000). Importantly, it is unclear what specific behaviors constitute sexual grooming, given that the behaviors may not be unlike normal adult/child interactions (Craven et al., 2006), and there has yet to be a validated model of the sexual grooming process. The lack of a comprehensive understanding of sexual grooming produces confusion amongst clinicians, law enforcement, attorneys, researchers, and community members alike. As such, the present study sought to establish content validity of a sexual grooming model, including both the stages and specific behaviors that are involved in the process. Sexual grooming Sexual grooming has become synonymous with CSA in the past several decades (McAlinden, 2013). The goals of grooming are to gain initial coopera- tion of the victim, decrease the likelihood of discovery, and increase the likelihood of future sexual contact (Lanning & Dietz, 2014). These pre- offense behaviors are thought to be a deliberate process that is highly complex and nuanced, with behaviors often mirroring normal adult/child interactions (Knoll, 2010; McAlinden, 2013). Therefore, it is difficult to establish represen- tative prevalence rates of the number of child sexual abusers who employ sexual grooming tactics in the offense process. Of the few studies that have tackled this question, it is estimated between 30 to 45% of child sexual abusers groom their victims (Canter et al., 1998; Groth & Birnbaum, 1978). Grooming can encompass varying behaviors which may differ based on the characteristics of the offender (e.g., age of the offender) and the victim (e.g., age or gender of the victim), as well as contextual factors (e.g., "effectiveness" of the grooming tactics, the offender's relationship to the victim, cultural factors; Kaufman et al., 2006). Notably, sexual grooming can occur both in- person or online. Online and off-line grooming processes may differ in important ways, as there are some behaviors that are not possible online (e.g., providing the victim with alcohol; Elliott, 2017) and thus, the present study will focus solely on in-person grooming behaviors. It should also be noted that individuals who sexually abuse children may groom themselves (personal grooming) and other people (familial and institu- tional/community grooming), in addition to the child. Personal grooming involves the process whereby the offender grooms themselves in order to 3502-033 Page 3 of 22 EFIA_00001665 EFTA00157006  JOURNAL OF CHILD SEXUAL ABUSE 0 3 Table 1. Sexual grooming model. Victim Selection = I-CVI Compkant/trusting of adults 0.78' Lacks confrlencelow self-esteem 089' Lonely/isolated 0.78' Troubled 089' Needy 089' Umvantedlunloved 089' Not dose to parents/parents are not 0.78' resources for them Single mothers/need of lather figure' 089' Lack of supervision 0.94' Gaining Access and Isolation (a = 5) Involvement in youth-serving organizations 083' Manipulate family to gain access to child 1.00' Activities alone with children/excludes adults 089' Overnight stays/outings 0.94* Separate child from peers and family 089' Trust Development la = 10) Charming/nice/likable 083' insider status/good reputaucorpillar 0.78' of the community' Affectionate/lovig LOW Giving the child attention 1.00' Favoritismrspecal relationship" 089' Compliments 089' Spending time with child/communicating 0.94' often Engage in oNldlike activities (e.g., stories. 089' games, sports, music) Rewards/privdeges (e.g., gifts, toys, treats. 0.94' money, tics) Provided drugs and/or akohol 089' Desensltlntlon to Sexual Content and Physical Contact (a = 10) Ask questions about child's sexual 089' experience/relationsNps Talk about sexual things they themselves 0.94' had done Inappropriate sexual language/dirty Jokes 083' Teach child sexual education 089' Use of accidental toucNrig/distracnon 089' wive touching Watch the child undreuing 0.78' Exposing naked body 0.78' Show child pornography magazines/videos 083' Seemingly innocent/non-sexual contact 0.94' Desensitize to touchencreasing 1.00' sexual touching Post-Abuse Maintenance Behaviors 0 = II) Told not to tell anyone what happened 0.89' Encouraging secrets 0.89' I love you/you're special 100* Rewards/bribes/avoid punishment 089' Persuaded the child it was 089' acceptable/normal behavior Misstated moral standards regarding touch 083' Victim made to feel responsible 0.78' Threats of abandonment/rejection/family 083' breaking up Items Not Included in the Five-Stage Model 0 = 35) (Continued) 3502-033 Page 4 of 22 EFTA_00001666 EFTA00157007 4 ® G. At. WINTERS ET AL. Table 1. (Continued). Victim Selection In = 9) I-CVI Selects a child who has already been victimized 0.72 Selects a child who is depressed/unhappy 0.72 Talks to the child on their level 0.72 Say things about the child's body/dress 0.72 Goes into child's bedroom while the child is in there 0.72 Goes into the bathroom while child is in there 0.72 After the abuse, threatens victim 0.72 Gains access after being approached by a chlidThad a chid recruit 047 other children Treats the child like an adult 047 Engages in verbal threatefinghten/intimidate/coercion of the child 047 Violates the child's privacy 047 Has the child observe sexual behavior 047 Selects a child who is cognitively impaired/special needs/learning 041 disability Selects a child who has drug or alcohol abusing parents 041 Looks at/inspects child's body for development 041 Selects a child who has economic problems/parents working a lot 036 Gives the child rides home 036 Babysits the child 036 Gains access to children through public places (e.g., malls, arcades) 036 Takes photos/videos of the did 036 Selects a child who is young or small/slim 030 Selects a child who parents are divorced/marital problems 030 Selects a child who has a mother who was sexually abused 030 Uses stre/authority/streregth against the add 030 Selects a child who is attractive/pretty (e.g. hair type, skin color) 0.44 Shows helpfulneu to others 0.44 Looks at did in a funny/sexual way 018 After the abuse, the offender assumes the child's silence 028 Selects a child based on his/her clothing 032 Has the child view violence against others 032 After the abuse, the offender punishes the child 032 Punishes the child or withholds privileges 0.17 Use of physical force/uses weapons against the child (e.g., push, 0.17 shove, spank) Presents as mean/rude to the child 0.11 After the abuse, the offender moves on to the next victim 0.11 ' indicates significant results justify, minimize, or deny their behaviors (Craven et al., 2006; McAlinden, 2006). The purpose of familial grooming is to gain the trust of caregivers in order to increase access to the victim and decrease the likelihood of disclosure. An offender may also engage in community or institutional grooming, such as becoming a respected member of society or seeking careers or volunteer positions that allow access to children (e.g., Boy Scouts, schools, foster care; McAlinden, 2006; Sullivan & Beech, 2002; Van Dam, 2001). Given the complicated nature of identifying sexual grooming, it has been proposed that recognizing sexual grooming behaviors following the disclosure of the sexual offense is much easier than prospective identification (Craven et al., 2006). Researchers have found there is a hindsight bias associated with sexual grooming of children, in which individuals tend to overestimate the likelihood that they could have predicted these behaviors were taking place 3502-033 Page 5 of 22 EFIA_00001667 EFTA00157008  JOURNAL OF CHILD SEXUAL ABUSE ® S after they learn an individual has committed a sexual offense (Winters & Jeglic, 2016). Importantly, in one study, Winters and Jeglic (2017) found that the general public has trouble identifying potentially predatory sexual grooming behaviors. Given the difficulty in identifying sexually versus non-sexually driven behaviors with children, gaining a better understanding of sexual grooming is integral to improved prevention and treatment efforts. Legal definition of sexual grooming It should be noted that the legal definition of sexual grooming is not necessa- rily synonymous with concept of in-person sexual grooming as outlined in the scientific and theoretical literature. By 2017, 63 countries had enacted legisla- tion related to grooming that focuses solely on the online solicitation of minors (often referred to as online sexual grooming; International Centre for Missing and Exploited Children, 2017). Notably, many of these laws do not account for sexual grooming that can occur in-person. Other countries have developed legislation that could be applied both to online and in-person grooming cases. For example, in the United States, section §2422 of the federal Criminal Code describes a law whereby an individual who "knowingly per- suades, induces, entices, or coerces any individual to travel in interstate or foreign commerce, or in any Territory or Possession of the United States, to engage in prostitution, or in any sexual activity for which any person can be charged with a criminal offense, or attempts to do so" can be fined or imprisoned (Coercion and Enticement, 18 U.S.C. 2422). While the aforemen- tioned law pertains particularly to cases involving sex trafficking, several states have followed suit and enacted similar laws without the requirement of "interstate or foreign commerce" which can then more generally apply to cases of CSA involving grooming. It is important to have a legal definition of sexual grooming for the purposes of prosecution of these crimes; however, legal definitions typically lack specificity (e.g., what behaviors that would be indicative of grooming). Further, and most importantly, in order to prevent grooming-based CSA, it is vital to go beyond the legal definitions to better understand the interaction between the victim, offender, and context of the offense (e.g., Nash & Williams, 2008). Thus, the grooming behaviors analyzed within this paper will be clearly differentiated from that of the already accepted legal definitions. Models of sexual grooming There have been numerous attempts to identify the steps involved in the sexual grooming process and to develop an overarching model of these behaviors (see Appendix A); however, none of these models have been empirically validated. One of the most widely cited models of sexual grooming authored by 3502-033 Page 6 of 22 EFTA_00001668 EFTA00157009 6 G.M. VANTERS ET AL. McAlinden (2006) indicates, as described above, that offenders groom not only children, but also themselves (i.e., personal grooming) and family and community members who act as gatekeepers to the children. Another widely cited grooming framework by Elliott (2017) - the Self-Regulation Model - draws upon the strengths and limitations of previous models of grooming. The model is comprised of two phases: 1) the potentiality phase includes rapport building, incentivization, disinhibition, and security management; and 2) the disclosure phase which describes how gains made in the first phase enable the perpetrator to desensitize the victim to sexual abuse. Although the self- regulation model of sexual grooming advanced the field, this model is not easily understood or applied, and thus, a more simplified model is greatly needed to enhance communication across fields. In an effort to address some of the limitations of previous models of grooming behavior, Winters and Jeglic (2017) reviewed the extant grooming literature and developed a model of grooming comprised of behaviors that could be observable to others and measurable, and thus informative in pre- vention and detection of sexual abuse. This five-stage model, hereafter referred to as the Sexual Grooming Modal (SGM), draws upon the commonalities identified in several of the previously proposed models (see Appendix A), as well as identifying gaps of missing information. For example, some previously proposed models did not address important components of grooming, such as victim selection or post-abuse maintenance (e.g., Brackenridge, 2001; Sheldon & Howitt, 2007). Additionally, other models have limited utility for public prevention initiatives as they are theoretically complex and thus difficult to apply in real-world settings (e.g., Elliott, 2017; Olson et al., 2007). Winters and Jeglic (2017) model of grooming behavior proposes five overarching stages that may be involved in the complex process of sexual grooming, including: 1) selecting a victim; 2) gaining access and isolating the victim; 3) developing trust with the child and others (e.g., caretakers, community members); 4) desensitizing the child to sexual content and physical touch; and 5) main- tenance behaviors following the commission of the abuse. Below, each stage is described with support from the theoretical literature. Victim selection First, several models of grooming propose that selecting a vulnerable victim is the initial step in the grooming process (e.g., Harms & van Dam, 1992; Lanning, 2010). It has been proposed that a vulnerable child may be identified based on physical characteristics (e.g., child who is perceived as attractive, young, or small; Conte et al., 1989; Elliott et al., 1995), or emotional or psychological needs (e.g., child who is perceived as trusting, lacking self- esteem, isolative, neglected, troubled, or in need of affection; Elliott et al., 1995; Kaufman et al., 2006; Knoll, 2010; Shakeshaft, 2004). Additionally, an 3502-033 Page 7 of 22 EFTA_00001669 EFTA00157010  JOURNAL OF CHILD SEXUAL ABUSE 0 7 offender may look to the child's family circumstances in the victim selection process (e.g., lack parental supervision, parental discord, parental mental health/substance use issues; Craven et al., 2006; Jackson et al., 2015; Kaufman et al., 2006). Gaining access and isolation Second, many of the prior models identify that an offender seeks to gain access to the targeted child and isolate him/her from others. Indeed, Lanning (2010), Craven et al. (2006), Olson et al. (2007), and Leclerc et al. (2009) all proposed models that include a stage whereby an offender gains access to the victim. Gaining access to a potential victim may include becoming involved in youth- serving organizations (e.g., Lanning & Dietz, 2014), frequenting public places with children (e.g., Kaufman et al., 2006), or manipulating the family in order to gain access to the child (e.g., Knoll, 2010; Lanning & Dietz, 2014). Once an offender has gained access to a child, they often work to isolate the child physically and emotionally from their family and peers (e.g., Craven et aL, 2006; Lawson, 2003). For example, an offender may seek to organize activities that physically isolate the child all the while excluding adult involvement, such as overnight stays, giving the child a ride home, or babysitting the child (e.g, Kaufman et al., 2006). Trust development Third, after selecting and gaining access to a victim, prior models describe a stage in which the offender works toward deceptively developing trust and cooperation with the child (Craven et aL, 2006; Leclerc et at, 2009; Olson et al., 2007). While some models incorporate a broad stage that refers to the overarching goal of trust development, others have outlined specific behaviors that may be used to gain the trust. An offender may try to present as likable and charming, eventually earning insider status and a good reputation in the community (e.g., Lanning & Dietz, 2014). The offender may make the child feel loved, use bribes or inducements, exploit his/her vulnerabilities, engage in peer-like activities, and befriend the child (Berliner & Conte, 1990; Harms & van Dam, 1992; Lederc et al., 2009; Marshall et al., 2015). Additionally, literature has identified that some offenders may provide the child with drugs or alcohol (e.g., Bennett & O'Donohue, 2014), which would be most commonly used with older victims. Desensitizing the child to sexual content and physical contact Fourth, there appears to be a stage that involves the introduction of sexual conversation and touch, with the aim of desensitizing the child to these behaviors (Berliner & Conte, 1990; Harms & van Dam, 1992; McAlinden, 2006; Olson et al., 2007). An offender may introduce sexualized topics into 3502-033 Page 8 of 22 EFTA_00001670 EFTA00157011 8 G. M. VANTERS ET AL. discussions, such as telling inappropriate jokes, providing sexual education, or engaging in sexual conversations (Knoll, 2010; McAlinden, 2006; Olson et al., 2007; Wyre, 2000). The offender may violate the child's privacy (e.g., spying, sneaking views of the child; Bennett & O'Donohue, 2014) or engage accidental touching (Harms & van Dam, 1992; Olson et al., 2007). Moreover, literature commonly refers to a process by which an offender desensitizes the child to touch by gradually increasing physical contact (Berliner & Conte, 1990; Harms & van Dam, 1992; McAlinden, 2006). For example, the individual may begin using tactics such as hugging or tickling, then gradually increasing contact over time to wrestling or massages. Post-abuse maintenance Finally, an offender may engage in maintenance behaviors which are used to continue ongoing abuse with the victim and/or prevent disclosure (e.g., Craven et al., 2006; Harms & van Dam, 1992). It has been suggested that this stage involves the offender encouraging the child to maintain secrets and not disclose the abuse (Craven et al., 2006; Harms & van Dam, 1992). An offender may try to persuade the child that the sexually abusive behavior is acceptable (e.g., Jackson et al., 2015), misrepresent standards for appropriate touching (e.g., Bennett & O'Donohue, 2014), or make the child feel respon- sible for the abuse (e.g., Harms & van Dam, 1992). Affection may also be employed by telling the child they love them or the child is special (Lang & Frenzel, 1988), giving the child bribes or rewards (e.g., Lang & Frenzel, 1988; Lawson, 2003; Salter, 1995; Shakeshaft, 2004), or enforcing or withholding punishment (Lawson, 2003). While Winters and Jeglic (2017) SGM addresses the limitations of previous models, similar to all the other past models of sexual grooming, this model has not yet to be validated. Given that isolated grooming-like behaviors in and of themselves may not be indicative of sexual abuse, it is necessary to establish a model of the stages of grooming to understand the larger process in order to inform detection and prevention efforts. Thus, the present study aimed to be the first to empirically validate a model of sexual grooming and identify what specific behaviors constitute grooming. The present study The present study aimed to establish the content validity of the proposed SGM (Winters & Jeglic, 2017) and identify what behaviors may be indicative of sexual grooming. To this end, experts in the field were asked to identify whether they believed the five stages of the SGM are part of the grooming process and what specific behaviors (identified from a thorough literature review) might fall under each of these proposed stages. Given the lack of 3502-033 Page 9 of 22 EFTA_00001671 EFTA00157012  JOURNAL OF CHILD SEXUAL ABUSE 0 9 previous empirical research in this area, the study was exploratory in nature and thus, no specific hypotheses were made. Method Part 1 Literature review A comprehensive literature review was conducted to identify potential groom- ing tactics that have been identified in previous publications. Online searches for articles were conducted through Psycl NFO, Criminal Justice abstracts with Full Text, Web of Science, and Medline Complete. The search terms utilized included: I) Set' groom* and 2) Child* groom*. Sources were also found by reviewing the reference lists of sources obtained through these online database searches. Searches were limited to articles in English language and peer- reviewed sources. A total of 1,363 sources resulted from literature search of the four search engines and reference lists. These sources were screened using a review of titles and abstracts, which resulted in the collection of 69 initial sources. Following a full-text review of the sources, 51 articles and books were identified as relevant. These sources all contained information regarding sexual grooming behaviors enacted by in-person child sexual abusers (i.e., online sexual grooming literature was excluded). The 51 articles and books were thoroughly reviewed, and each unique grooming behaviors was recorded in order to produce a comprehensive list of possible grooming behaviors. Through this process, a total of 77 potential grooming behaviors were identified.' Part 2 Participants and procedures Content validity of the five-stage SGM and 77 grooming behaviors was examined by having a list of "experts" in the field complete an online survey. The list of experts was developed by compiling a list of authors (ii = 99) on the articles and books that were published in the area of sexual grooming (described above). Extensive research was conducted through the use of Internet search engines and contact information listed within the literature to identify the e-mail addresses of the authors. A total of 56 e-mail addresses were obtained. Following sending e-mail invitation to participate in the study, 12 e-mail addresses were inactive which resulted in a total of 44 potential participants. Three rounds of e-mails were sent to each e-mail address 'n should be noted that the authors also seated an a piton model which identified which of five stages each behavior fell under, lit was later utikzed in making final determinations regarding what stage of the grooming process each relevant (as identified by experts in the field) grooming behavior would likely be utilized. 3502-033 Page 10 of 22 EFTA_00001672 EFTA00157013 ICI 0 G. M. WINTERS ET AL requesting participation in the expert review which involved participation in a 30-minute survey. If the individual agreed to participate, they were asked to complete the Expert Review Survey (see below). A total of 18 participants completed the survey (12 males; 6 females), which represented a 40.9% response rate. In regard to participant age, four indivi- duals were between the ages of 41-50, six between the ages of 51-60, and eight over the age of 60. The majority of experts obtained a Ph.D. (ii = 15), two had a Master's degree, and one was a current Ph.D. student. There was a range of fields in which these degrees were earned: psychology (n = 8), criminal justice (n = 2), and one individual each from the fields of education, sociology, public health, social work, communication, theology, criminology, and psychology/ sociology. Experts reported the area, or areas (respondents could select more than one), that best described their experience working with child sexual abusers, which included empirical research (n = 16), clinical practice (ii = 10), publishing theoretical articles/chapters on the topic (n = 14), and other (n = 3; i.e., employee of state correctional system, consultation on investigations, investigative journalist). The experts reported a mean number of years of experience with empirical research (ii = 18), publishing theoretical pieces (n = 17), and clinical experience (n = 10) related to grooming as 24.67 years (range = 2-48), 15.50 years (range = 3-48), and 22.71 years (range = 2-45) years, respectively. All participants (n = 18) had published an empirical research article related to grooming, with 12 individuals reporting between 1-10 publications, two reporting 11-20 publications, three with more than 20 publications, and one participant indicated that they were not certain how many publications they had. For the 17 people who had experience publishing theoretical articles/chapters on sexual grooming, the mean number of publications was 7.00 (range = 1-20). Of the 10 participants who had clinical experience with sex offenders, six had 50 or more clients, two had 15-50 clients, one had 5-15 clients, and one had 0-5 clients. Expert review survey First, participants were presented with 4-point Likert scale items inquiring about the relevance (1 = not relevant, 2 = somewhat relevant, 3 = relevant, 4 = very relevant) of the five proposed stages of grooming. Second, the participants rated the relevance of each item from the pool of 77 grooming behaviors identified by the literature review using a 4-point Likert scale (1 = not relevant, 2 = somewhat relevant, 3 = relevant, 4 = very relevant). Participants were also asked for each item to select one or more stages of the grooming process the behavior fell under (i.e., Victim Selection, Gaining Access, Trust Development, Desensitization, and Post-Abuse Maintenance, other, or none). Lastly, participants completed a series of demographic ques- tions (e.g., age, gender, degree, field of study, clinical, publication, and research experiences). 3502-033 Page!! of 22 EFTA_00001673 EFTA00157014  JOURNAL OF CHILD SEXUAL ABUSE 0 11 Results Analytic strategy The Content Validity Index (CVI) is a method originally proposed by Lynn (1986), which utilizes feedback from experts in the field to determine what content is relevant to a construct; this is a commonly used method in social science research (Research Methods Knowledge Base, n.d.). In this case, CVI calculations were used to determine what stages and behaviors are relevant to the process of sexual grooming. First, as noted above, the relevance of the five stages and potential grooming behaviors were rated by experts using a 4-point Likert scale (1 = not relevant, 2 = somewhat relevant, 3 = relevant, 4 = very relevant). Second, these ratings were used to determine which stages/behaviors should be retained (i.e., they were deemed related to the construct of grooming by the experts) and which should be rejected (i.e., they were deemed not related to the construct of grooming by the experts) through the utilization of CVls, which are calculations that examine the proportion of experts who rated the item as relevant. The CVI for each item (I-CVI) is calculated by dividing the number of experts who believed the item was relevant (either a 3 or 4 on the Likert scale) by the total number of content experts (in this case, n = 18). It has been suggested that the I-CVI for an item should be greater or equal to 0.78 in order to be included (Shi et al., 2012). Stages of sexual grooming Experts were asked to rate the relevance for each of the five stages of the sexual grooming process as proposed by Winters and Jeglic (2017). Results revealed an I-CVI index of.94 (17/18 experts) for the stages of Gaining Access (M = 3.78, SD = .55), Trust Development (M = 3.72, SD = .58), and Desensitization (M = 3.50, SD = £2). Similarly, an I-CVI index of .89 (16/18 experts) was found for the stages of Victim Selection (M = 3.56, SD = .70) and Post-Abuse Maintenance (M = 3.39, SD = .70). Overall, the I-CVls for each of the proposed stages exceeded the cutoff score of 0.78, suggesting that all five stages are believed to be relevant to the sexual grooming process. Sexual grooming behaviors An examination of the I-CVls for the 77 potential grooming behaviors revealed that 42 items were considered by the expert panel as relevant to the construct of sexual grooming (I-CVIs ranged between .78-1.0; see Table 1). This represents a retention rate of 54.5% from the original items. An examination of which stage of the grooming process the experts believed the behavior belonged in was conducted. For each item that was deemed relevant (n = 42), the stage that the most experts (i.e., over 50%) believed the 3502-033 Page 12 of 22 EFIA_00001674 EFTA00157015 12 0 G. M. WINTERS ET AL behaviors to fall under was recorded. These expert-rated categorizations were compared to the theoretical categorization identified by the researchers (see footnote on page 11). Results suggested that 39 of the 42 relevant items were deemed by the majority of the experts to fall into the original a priori model developed by the authors. One item ("Threatens the child with abandonment/ rejection/family breaking up") was rated by the majority of participants = 14) to fall under the Post-Abuse Maintenance stage, not the theoretically suggested Trust Development stage. Given the agreement among the vast majority of experts, this item was relocated to the Post-Abuse Maintenance stage. Two items ("Becomes involved in activities alone with children/excludes adults" and "Presents as charming/nice/likable to others") were rated by the experts as equally belonging to the Gaining Access and Trust Development stages. Consistent with the theoretical literature and a priori model, these items were deemed to fall under the Gaining Access and Trust Development stages, respectively. See Table 1 for the final grooming behaviors organized into the five-stages of the SGM. Discussion The present study aimed to establish content validity for the SGM proposed by Winters and Jeglic (2017) and identify which behaviors are involved in each stage of the grooming process. The results, as determined experts in the field, revealed consensus that the five stages proposed by Winters and Jeglic (i.e., Victim Selection, Gaining Access, Trust Development, Desensitization, and Post-Abuse Maintenance) are all essential components of the sexual grooming process. Moreover, findings from the study suggest there are 42 grooming tactics/behaviors that experts identified as belonging to these stages. Overall, the results of the present study resulted in the content validation of a comprehensive and parsimonious model of sexual grooming. Stages of sexual grooming A major benefit of the SGM's framework is that it is intuitive, easily under- stood, and backed by a foundation of literature. Having a comprehensive, yet easily understood, framework is vital, as information about grooming must be distributed to various consumers (e.g., law makers, researchers, parents, clin- icians, criminal justice professionals). For example, parents could utilize this model to be vigilant in monitoring for potentially predatory behaviors of those around their child, while clinicians can use the model to assist in assessment and treatment of victims or offenders of CSA. As such, it is important to have a model of grooming that allows for education across fields and different types of consumers. 3502-033 Page 13 of 22 EFTA_00001675 EFTA00157016  JOURNAL Of CHILD SEXUAL ABUSE 0 13 While establishing the content validity of the SGM is a major advance in understanding grooming behaviors, it remains but a first step. With the foundation provided by the findings of the current study, it is necessary to continue to establish empirical support for the model and begin to assess other facets of grooming behavior. For example, it is unknown whether every offender progresses through each of the five stages, or whether there is always a linear progression through the stages. For instance, if an offender already has preexisting access to the potential victim (e.g., a parent), then they are less likely to employ behaviors in the Victim Selection or Gaining Access stages. Moreover, it may be that the offender moves fluidly between stages or skips stages if not deemed necessary. As an example, if an offender utilizes behaviors in the Desensitization stage and then notices the child resisting, they may revert back to engaging in more behaviors in the Trust Development stage. Similarly, the proposed model does not assume that an offender may only utilize behavior within one stage at a given time; that is, an offender may simultaneously employ behaviors found in the Trust Development (e.g., show- ing the child affection) and Desensitization (e.g., using seemingly innocent touch) stages. Taken together, future research should aim to examine the types, and most common, progression of the stages during the offense process. Sexual grooming behaviors Overall, the study was the first to obtain data related to relevance of various behaviors to the grooming process. This is an important addition to the litera- ture given that it has previously been unclear what behaviors constitute groom- ing, especially given that many grooming behaviors in and of themselves are not unlike normal adult/child interactions. Identifying the 42 behaviors that were deemed relevant to the sexual grooming process by experts in the field is an important advance. While the data has yet to be empirically validated using cases of CSA, an expert-review validation study is the first step in better understanding what behaviors are indicative of grooming. It should be noted, however, that we did not ask experts to provide items that they believed to be indicative of the stages of sexual grooming. Rather, the items were provided to them to endorse. This could lead to a reification effect in that that the experts may have endorsed items as relevant to the stages of grooming given the items were derived from existing theoretical grooming literature, yet the items they endorsed may not in fact represent concrete behaviors actually utilized by perpetrators in CSA cases. However, if that were the case, then the majority of items would have been retained as relevant in the study as they were extracted from the grooming literature, when in the study we found that only about half of the theoretically- linked items were deemed not to be indicative of grooming. Thus, it is likely that the experts were critically evaluating the items to determine which were applic- able to real-world cases. 3502-033 Page 14 of 22 EFTA_00001676 EFTA00157017 14 0 a M. WINTERS ET AL Taken together, a major strength of the SGM is that the behaviors that are observable and measurable, although it remains unclear how to differentiate these behaviors from innocent contact with children. Nonetheless, we have garnered a greater understanding, using expert consensus, of actions that may be employed by a would-be child sexual abuser. That is, a validated model will assist in identifying constellations of behaviors that are considered grooming, which is a necessary component of preventing CSA. Moreover, the SGM provides a framework for the development of an instrument that can be used to measure sexual grooming, which can help identify and quantitatively measure the likelihood that a constellation of behaviors constitutes grooming. Implications of the sexual grooming model Overall, the results of the study have imp