Stages of Sexual Grooming: Recognizing Potentially Predatory Behaviors of Child Molesters

ABSTRACT

It has been proposed that sexual grooming behaviors of child molesters are not easily identified prior to the abuse. The present study investigated whether participants could recognize grooming behaviors, and if so, identify which stages of the process were most easily identified. Participants were randomly assigned to read one of six vignettes describing the stages of the grooming process. Results revealed no differences in likelihood ratings that the person in the vignette was a child molester and would commit sexual abuse between any of the conditions, suggesting that people may be unable to identify potentially predatory behaviors of child molesters prospectively.

Some types of sexual offending require an offender to use charm and manipulation to coerce a victim into a sexual relationship, a process that may take place over an extended period of time (Sasse 2005; Wood, Wilson, and Thorne 2015). Empirical research has found that nearly half of the offenders who commit sexual acts against children utilize what are known as “grooming” behaviors (Canter, Hughes, and Kirby 1998). Grooming refers to the behaviors that an offender employs in preparation for committing sexual abuse against a child (McAlinden 2006). While several definitions of grooming have been suggested, one of the most specific and comprehensive explanations was proposed by Craven, Brown, and Gilchrist (2006:297), who defined sexual grooming as:

A process by which a person prepares a child, significant others, and the environment for the abuse of this child. Specific goals include gaining access to the child, gaining the child’s compliance, and maintaining the child’s secrecy to avoid disclosure. This process serves to strengthen the offender’s abusive pattern, as it may be used as a means of justifying or denying their actions.

Successful grooming involves the skillful manipulation of a child and the community so that sexual abuse can be more easily committed without detection (van Dam 2006). Child molesters who groom often do so while keeping a guise of being kind, charming, and helpful (van Dam 2001). They strategically manipulate the victim, their family, and the community to hide their deviant intentions and avoid detection. These behaviors include strategies such as selecting a vulnerable victim, gaining access to the child, developing trust, and desensitizing the victim to touch.

A vast amount of research has supported the notion that many cases of child sexual abuse are never disclosed (Salter 2003) and it is speculated that as few as 5% of child sex offenders that are ever apprehended. Often after a child molester is apprehended there is outrage as the public questions why none of the abuser’s behavior was detected beforehand. However, Gillespie (2002) and Ost (2004) have suggested that it is highly unlikely that police and community members will be able to detect grooming behaviors that occur before the abuse in cases of intrafamilial and professional child abuse. Additionally, Craven and colleagues (2006) proposed that identifying grooming behaviors is more easily done retrospectively, compared to prospective identification. This is likely due to the fact that many grooming behaviors appear to be innocent in nature and typical of adult child interactions, while the motivation behind the behaviors is sexually deviant (Craven et al. 2006). In addition, offenders will change their strategies if they fear disclosure, making them even more difficult to identify (Conte, Wolf, and Smith 1989). Distinguishing between sexually motivated grooming and normal adult/child interactions is especially difficult when the witnesses have no knowledge of the grooming tactics employed by child molesters. Thus, there is a significant need to learn more about the sexual grooming behaviors of these offenders and how they may be identified prior to the commission of the abuse.

To our knowledge, only one study has empirically explored people’s ability to recognize sexual grooming behaviors. Using vignettes of sexual grooming scenarios, Winters and Jeglic (2016) found participants rated the likelihood of a person being a child molester significantly higher when the grooming behaviors were present in a vignette than when they were not. Additionally, the behaviors that participants were most able to identify as potentially inappropriate were those involving physical touch and isolation. Thus, the present study seeks to expand on the empirical grooming literature by exploring whether certain stages of the grooming process are more easily identified. Gaining a better understanding of individuals’ knowledge of sexual grooming tactics can provide valuable information toward early recognition and prevention of child sexual abuse.

The stages of the grooming process

The sexual grooming literature suggests that the grooming process consists of a series of stages that the child molester progresses through as they groom the victim for the impending sexual abuse (Lanning 2010; Leclerc, Proulx, and Beauregard 2009; McAlinden 2006). It must be noted that, to date, there has yet to be a validated model of sexual grooming. It has been argued that it is difficult identify the nature and extent of the grooming process (Williams 2015). As has been found for online sexual grooming behaviors, there may be a high level of variance within the grooming process depending on the individuals involved (Whittle, Hamilton-Giachritsis, and Beech 2015). Further, it may be near impossible to precisely pinpoint where the sexual grooming process begins and ends or the fluidity of the behaviors throughout the offending process (Mooney and Ost 2013). However, there has been a vast amount of literature that suggests there are common grooming behaviors that occur during the processes of selecting a vulnerable victim, gaining access to the child, developing trust, and desensitizing the victim to touch. Thus, the present study refers to “stages” or “steps” as clusters of grooming behaviors that some predatory child molesters may enact.

The first step in grooming is the selection of a victim, which can be based on appeal/attractiveness, ease of access, or perceived vulnerabilities of the child (Lanning 2010; McAlinden 2006; Mooney and Ost 2013; Olson, Ellevold, and Rogers 2007). Elliott, Browne, and Kilcoyne (1995) found that when selecting a victim, offenders often focused on physical characteristics such as the child being pretty (42%), the way the child is dressed (27%), or the child being small (18%). In addition, child molesters are also found to target children based on family situations, such as those living in single family households as often in these cases children may have less adult supervision or the custodial parent may rely upon others to help with childcare responsibilities (Conte et al. 1989; Elliott et al. 1995; Lang and Frenzel 1988; Olson et al. 2007). Further, children who have families with alcohol or drug addictions, emotional or mental problems, marital discord, domestic violence issues, or that are neglectful are at higher risk for sexual abuse as these situations may also lead to less parental supervision (Olson et al. 2007). The offender may also seek children with perceived psychological vulnerabilities that would allow the child to be more easily isolated from others, such as low self-esteem, low confidence, insecurity, neediness, or naivety (Finkelhor 1994; Olson et al. 2007). Elliott and colleagues (1995) found that 49% of child molesters targeted children who lacked confidence and self-esteem and another 13% targeted innocent and trusting children. Within the online grooming literature, it has been proposed that a child who is isolated and lacks social support may be more likely to engage with a stranger who is offering acceptance (Williams, Elliott, and Beech 2013), which is likely the case for in-person grooming as well. The research has shown that victim selection is an incredibly strategic, well-planned process that marks the first step in the sexual grooming process.

The second stage of the grooming process involves the child molester gaining access to the potential victim, with the goal of isolating the child both physically and emotionally from those around them (Lanning 2010; Olson et al. 2007). Intrafamilial offenders, those who commit offenses against a family member, by nature of the preexisting relationship to the victim, have victims who are easily accessible generally in the home environment. For example, Lang and Frenzel (1988) found that 41% of incest offenders will sneak into the child’s bedroom. Extrafamilial offenders, on the other hand, often seek out situations where victims will be readily available, such as going to malls, arcades, schools, parks, pools, and fairs (Elliott et al. 1995). Additionally, Elliott and colleagues (1995) found that 33% of extrafamilial offenders work on being welcome in the victim’s home as a means of getting closer to the children. These types of offenders may offer to help out around the house, assume a “father figure” role for the child, offer to babysit, or suggest overnight stays (Lanning 2010). Child molesters often seek jobs that involve contact with children like teachers, camp counselors, bus drivers, or coaches. Those who establish themselves in professional settings may create reasons to see the child after school hours or offer to take them on outings. These activities typically exclude other adults, in an attempt to get the child alone. Furthermore, in a model of grooming proposed by Mooney and Ost (2013), this second stage may involve engaging the child in communication (e.g., offering a ride, invitation to a party, offering drugs or alcohol), in order to initiate contact with victims who they do not readily have access to. All of these techniques are used to gain access to the victim allowing grooming and the abuse to be more easily enacted.

The next step in the grooming process involves the emotional recruiting of the victim. Deceptive trust development is a child molesters’ “ability to cultivate relationships with potential victims and possibly their families that are intended to benefit the perpetrator’s own sexual interests” (Olson et al. 2007:240). This step is often regarded as the central role of the grooming process, wherein the offender establishes trust and cooperation with the victim (Salter 1995; van Dam 2001). The offender accomplishes this by befriending the child, by learning about his/her interests, being helpful, showering the child with gifts and attention, or sharing secrets (McAlinden 2006). These behaviors are used to give the child the impression there is a loving and exclusive relationship between them and the would-be offender (Mooney and Ost 2013). The perpetrator portrays himself as a nonthreatening individual with whom the child can talk and spend time with. During this step, the offender adjusts his strategies based on the age of the child he is targeting, the needs of the child, and the child’s perceived vulnerabilities (Lanning 2010). Depending on the child’s age this may include engaging in peer-like involvement with the child, such as playing games with younger children or talking about sexual matters with teenagers. The child molester may also use inducements (e.g., money, treats, gifts, fun trips) in order to develop a “special relationship.” The main goal of this step is to establish trust, which then allows the offender to control and manipulate the child into participating in sexual abuse.

After the child molester has acquired the trust of the child, they may gradually increase physical contact in order to desensitize the child to touch (Berliner and Conte 1990; Christiansen and Blake 1990; Leclerc et al. 2009; McAlinden 2006). The child molester escalates physical contact to prepare the child for the sexual contact that will occur during the impending abuse. Often times this begins with seemingly accidental touch or innocent behaviors, which then escalate to more intimate touching. For example, the child molester may first give the child hugs or pats on the back, and then gradually escalate to wrestling, tickling, or back massages and the eventual sexual contact. Other tactics include playing hide and seek in the dark, playing strip poker, swimming nude, drying a child off with a towel, massaging an injury, playing physical games, cuddling, or showing the child pornography (Lanning 2010; McAlinden 2006). The most common strategies to introduce touch used by incest offenders include cuddling, sneaking into the bedroom, wrestling, or using sex as a game (Lang and Frenzel 1988) whereas non-incest child molesters use tactics such as cuddling, offering a massage, and using sex as a game. Berliner and Conte (1990) suggest that desensitization is not only physical, but can also be psychological. For example, the child molester may begin discussing sexual content in an attempt to achieve increased sexualization. Through a review of literature, there appears to be a series of stages the offender utilizes in sexual grooming, including selecting a vulnerable victim, gaining access to the child, developing a trusting relationship, and desensitizing the child to touch.

The current study

Given that grooming may be utilized by nearly half of sexual offenders against children (Canter et al. 1998), it is imperative that we better understand a way to detect their predatory behaviors in an effort toward prevention. Thus, the present study seeks to explore whether individuals can recognize the sexual grooming behaviors of child molesters. It should be stressed that not all sex offenders who commit crimes against children engage in these grooming behaviors. Further, there are various forms of grooming that can be enacted that will not be covered in the present article (e.g., online grooming, grooming of institutions or victim’s families, intra-familial sexual groomers). The focus of this study is on child molesters who use grooming to commit extra-familial offenses, which has been referred to as “predatory pedophilia” (Williams 2015). The present research seeks to examine whether individuals notice the predatory grooming behaviors used to subtly develop a relationship with the victim using seemingly every day social interactions (McAlinden 2012).

Past literature has suggested that these covert behaviors are not easily identified before the abuse occurs (Craven et al. 2006), yet the only empirical study to date suggests that in fact participants were able to identify sexual grooming, in particular behaviors involving physical touch and isolation (Winters and Jeglic 2016). However, it has yet to be determined whether individuals are better able to recognize certain stages of the grooming process (i.e., victim selection, gaining access, trust development, and desensitization to touch). Should certain stages of the grooming process be less identifiable than others, this information can be invaluable in prevention efforts to recognize potential sexual predators prior to the abuse. Therefore, the primary goal of the present study is to further examine whether grooming behaviors are recognized, and if so, which stages of the grooming process are most recognized of being indicative of future sexual abuse. Based on the findings from Winters and Jeglic (2016), it is hypothesized that participants will be able to recognize sexual grooming behaviors, in particular the stages involving gaining access and physical touch as compared to the stages of victim selection and trust development. This research is a preliminary step in gaining a better understanding of perceptions of grooming behaviors of predatory sexual offenders.

Method

Participants

The final sample was comprised of 393 undergraduate students (114 males, 278 females) following the removal of individuals who failed the manipulation check (i.e., those who failed to correctly answer three basic questions about information provided in the vignettes). Participants were students from a large urban university who completed the study to partially fulfill course requirements. The mean age of participants was 20.43 (SD = 3.32), with ages ranging from 18 to 47 years. The sample was ethnically and racially diverse comprised of 182 Hispanic/Latino (46.3%), 63 white/Caucasian (16.8%), 65 black/African American (17.3%), 49 Asian (12.5%), 16 biracial (4.1%), and 18 identified as “other” (4.6%) individuals.

Materials and procedure

Sexual grooming vignettes

A random number generator was used to randomly assign each participant to one of six one-page, double-spaced vignettes. All of the vignettes described a teacher named John who coached a youth boys’ baseball team, with one of the players being an 8-year-old child named Robbie. The vignettes were created with reference to the vignettes from Winters and Jeglic (2016), in addition to a thorough review of the grooming literature. The vignettes described grooming behaviors commonly used by child molesters (Berliner and Conte 1990; Campbell 2009; Christiansen and Blake 1990; Craven et al. 2006; Elliott et al. 1995; Lang and Frenzel 1988; Lanning 2010; Lecrelec et al. 2009; McAlinden 2006; Olson et al. 2007; Salter 1995; van Dam 2001, 2006). Additionally, experts in the field of sex offender research reviewed the scenarios for realism and accuracy.

The six vignette conditions were as follows: (1) All Grooming Stages; (2) Victim Selection; (3) Gaining Access; (4) Trust Development; (5) Touch Desensitization; and (6) Non-Grooming. In the All Grooming Stages vignette, John utilized numerous tactics typically employed throughout the entire grooming process (i.e., selecting a vulnerable victim, gaining access, trust development, and desensitizing child to touch). Each paragraph of the vignette featured a different stage of sexual grooming. Some examples of grooming behaviors include: Robbie is a physically small boy who lacks confidence and comes from a divorced family (i.e., victim selection); John volunteers at organizations serving children and spends his spare time at places frequented by children (i.e., gaining access); John gives special attention to Robbie and buys gifts for only some players (i.e., trust development); and John horses around with the boys and gives some players hugs (i.e., desensitizing child to touch).

For the Non-Grooming condition, the story mirrored the All Grooming Stages vignette, but with all of the grooming behaviors removed. Some examples of non-grooming behaviors include: Robbie is a physically strong boy who has confidence and comes from a happily married family; John volunteers at organizations serving adults and spends his spare time with his family; John does not give special attention to Robbie and John buys equipment for the team; and John stops the boys from horsing around and waves goodbye to the players. For the other four conditions, only the paragraph featuring the grooming behaviors for a particular stage was included in the Non-Grooming vignette.

Likelihood ratings

Participants were asked to “Estimate the likelihood that the following statements are true about John” (0 = definitely not true, 100 = definitely true), which included 10 “filler” items (e.g., John is an alcoholic, John is a father, John is a domestic abuser). However, the answer of interest was the extent that “John is a Child Molester.” Participants were also asked, “Estimate the likelihood of the following scenarios happening in the future” (0 = definitely not true, 100 = definitely true). Again, the list contained an additional eleven “filler” items (e.g., John will win the lottery, John will get divorced, John will lead his baseball team to a winning season), but the rating of interest was “John will sexually abuse Robbie.” These two dependent variables will be referred to as Child Molester and Sexually Abuse.

Procedure

The study was conducted via an online survey site. Before beginning the study, the participants first read the informed consent and if they agreed to participate, they were asked to electronically sign and date the form, as well as print a copy for their records. Participants were told the purpose of the study was to examine individuals’ perceptions of a person based on a short description. Subsequently, those who consented were randomly assigned to one of the six story conditions. The participant was instructed to read the vignette and then answer the series of likelihood ratings. Last, the participants answered demographic questions (i.e., gender, age, ethnic/racial origin) and read a debriefing form which included contact information for the researchers and mental health services. The survey took approximately 30 minutes to complete, for which each participant received one research credit toward a course requirement.

Results

An examination of the data revealed the Kolmogorov-Smirnov test for normality was violated for two dependent variables, Child Molester (D = .327, p < .001) and Sexually Abuse (D = .309, p < .001). In instances when normality is violated, it has been acceptable practice to use non-parametric tests (i.e., those that do not assume the data is normally distributed), delete outlier data, or transform the data (Nimon, 2012). Therefore, the data was analyzed using the non-parametric Kruskal-Wallis H test, which has been determined to be an appropriate alternative to the parametric tests of Analyses of Variance (ANOVA) or Mulitvariate Analysis of Variance (MANOVA; Burke 2001).

A Kruskal-Wallis H test was conducted for each dependent variable (i.e., Child Molester; Sexually Abuse) in order to determine if there were differences between the six groups (i.e., All Grooming Stages, Victim Selection, Gaining Access, Trust Development, Touch Desensitization, Non-Grooming). The Kruskal-Wallis H test showed there were no statistically significant differences in Child Molester ratings between the different conditions, χ² (5, N = 393) = 4.44, p < .488. Similarly, the Kruskal-Wallis H test for the Sexual Abuse scores did not differ between groups, χ² (5, N = 393) = 3.47, p < .628, therefore no further tests were conducted. See Table 1 for the means and standard deviations.

Table 1. Means and standard deviations for condition by for the child molester and sexually abuse variable.

		Child Molester	Sexually Abuse
Condition	n	M(SD)	M(SD)
Non-Grooming	73	13.04(21.283)	7.79(12.966)
Victim Selection	71	8.14(15.866)	8.82(18.529)
Gaining Access	72	15.07(22.129)	13.49(21.246)
Trust Development	59	13.839(23.759)	14.86(25.186)
Desensitizing to Touch	57	14.33(24.016)	12.75(21.918)
All-Grooming	61	17.30(27.242)	18.05(27.074)
Total	393	13.49(22.459)	12.39(21.471)

Discussion

It has previously been proposed that sexual grooming behaviors are difficult to identify (Canter et al. 1998; Craven et al. 2006; Lanning 2010), likely due to the fact that many of these behaviors appear innocent and may be typical of normal adult/child interactions. The only study that has empirically examined this hypothesis found that participants, contrary to expectations, were able to recognize the sexual grooming behaviors of a child molester. Therefore, the present study sought to further explore the extent to which individuals could identify these potentially predatory behaviors. Specifically, we sought to investigate which stages of the grooming process were more easily recognized. We hypothesized that the sexual grooming stages involving gaining access and physical touch would be more easily recognized than the stages of victim selection and trust development. Overall, the results revealed that participants were unable to recognize sexual grooming behaviors for any of the stages of grooming. While the present study provided preliminary information on perceptions of sexual grooming behaviors enacted by predatory extra-familial child molesters, continued research in this area is greatly needed to help shed light on this complex process.

Our findings showed no significant differences in the likelihood ratings that the person in the vignette was a child molester or would sexually abuse the child in the story for those in the Non-Grooming condition (i.e., contained no grooming behaviors) and All Grooming condition (i.e., contained behaviors from all stages of the grooming process). Thus, participants were not able to identify the potentially predatory behaviors that child molesters may employ. Furthermore, there were no significant differences between any other conditions that featured only one stage of the grooming process. This suggests that certain stages of the grooming process were not more easily identified as predatory compared to others.

The findings from the present study contrast those of Winters and Jeglic (2016), who found that participants had higher likelihood ratings that the person was a sexual predator when the grooming behaviors were featured in the vignette compared to absent. These alternative findings may be due to the fact that the present study used a modified version of the vignettes used by Winters and Jeglic (2016). Some grooming behaviors were added to the scenario (e.g., John enjoys spending time at places frequented by children; John buys candy and gifts for some players) while others were removed (e.g., John gives Robbie a ride home; Robbie sits on John’s lap), which may have led to differences in participants’ ability to pick up on these behaviors. Given there is sparse empirical literature on identifiability of specific grooming behaviors, it is possible the omitted or additional behaviors in the present study lowered participants’ recognition of sexual grooming. However, Winters and Jeglic (2016) did report that the mean likelihood ratings the person was a sexual predator were toward the lower half of the scale (i.e., “definitely not true”), which they suggested indicated that while the results were statistically different between grooming and non-grooming conditions, participants were not overwhelmingly able to identify sexual grooming. Similarly, in the present study, although not statistically different, the all-grooming conditions had higher likelihood ratings compared to the non-grooming conditions. It may be that people detect some of these behaviors as potentially inappropriate, but not to the extent the behaviors are flagged as overtly worrisome. Taken together, the findings of these two empirical studies suggest that individuals may have difficulty identifying grooming behaviors, and even if they can identify them, the certainty of the predatory nature of the behaviors is still quite low.

While the present study revealed that participants may not be able to recognize sexual grooming as previous research found (Winters and Jeglic 2016); these findings are in line with past theoretical propositions suggesting that sexual grooming is difficult to recognize (Canter et al. 1998; Craven et al. 2006; Lanning 2010). Sexual grooming behaviors tend to, in themselves, not be overtly sexual in nature and may appear innocent, such as hugging a child, working or volunteering with children, or showing a child attention. These types of child molesters intentionally manipulate the environment around them in order to go more easily undetected (van Dam 2002). An offender may appear to be acting innocent in nature, while the behaviors may in fact have deviant intentions (Craven et al. 2006). Therefore, it is not surprising that individuals have difficulties recognizing sexual grooming as overtly worrisome, given they tend to be inconspicuous in nature. Interestingly, the sample used in the present study were students from a university for criminal justice, who may have had some prior knowledge of sexual offenders. It is even more intriguing and provides strength to the findings that even with some possible prior knowledge of offenders, grooming behaviors still are not easily identified.

The results suggest that individuals have difficulties not only identifying more covert grooming behaviors (e.g., those involved in selecting a vulnerable victim and trust development), but also overt behaviors (e.g., those involved in desensitizing the child to touch and gaining access to the child). We expected that individuals would recognize overt behaviors as potentially inappropriate, such as hugging the child, taking certain children out for ice cream, or driving the child home. It is likely that both overt and covert sexual grooming behaviors are more easily identified retrospectively, rather than prospectively, as proposed by Craven and colleagues (2006). Once identified, people may be outraged at how the sexual offender was not identified sooner. However, as Winters and Jeglic (2016) found, once people know a person committed a sexual offense against a child they overestimate the likelihood they would have predicted it. This hindsight bias phenomenon could lead to blame of the victim’s family or community for not preventing the abuse. Importantly, our findings showed that people may not be able to recognize potentially predatory behaviors prior to the commission of the abuse. Placing blame for not recognizing sexual grooming behaviors may be counterproductive. Rather, efforts should be focused on educating the community on spotting these behaviors prospectively as a means of preventing child sexual abuse.

Limitations

The present study has some limitations that should be addressed. First, the study utilized an undergraduate student sample which is not representative of all individuals who may come in contact with a potential child molester. To better understand the extent to which these findings are generalizable to the population, further studies should seek to explore the ability to recognize grooming behaviors in parents, school staff, and other community members who frequently interact with children. Further, more research should be done examining other types of grooming behaviors in various scenarios, such as school staff, boy scout leaders, or relatives. The vignettes used in the study featured grooming behaviors that have been documented in prior research. However, given that there has yet to be a clear and comprehensive definition of grooming and the exact behaviors that would constitute grooming, it is difficult to validate the measure. There is much need for more research on sexual grooming to better understand these types of predators and the behaviors they utilize in the offense process.

Implications

The present study found that individuals are not able to identify potentially predatory behaviors that a child molester may employ. It may be that people were not able to identify the child molester who groomed because they are not aware of sexual grooming or what classifies as a grooming behavior. Thus, better educating children, parents, and community members about possible predatory behaviors could be greatly beneficial. Education could be provided to adults in the community through a variety of domains, such as pamphlets at locations frequented by caregivers, providing information to teachers and school staff, public service announcements, or websites geared toward families or education providers. In particular, those in frequent contact with children (e.g., school officials, guidance counselors) should remain cognizant that potential predators may be targeting vulnerable children and pay particular attention to these individuals. Caregivers should also aim to inform their children about potential dangers in an age-appropriate manner. Providing people with a better understanding of the sexual grooming process and who the offenders are could aide in identifying these types of perpetrators prior to the commission of the abuse.

Additional information

Notes on contributors

Georgia M. Winters

GEORGIA M. WINTERS is a clinical psychology Ph.D. student at John Jay College of Criminal Justice and the Graduate Center, City University of New York. Her research focuses on sexual grooming behaviors of child molesters and how sexual offenders are viewed by jurors in the criminal court system.

Elizabeth L. Jeglic

ELIZABETH L. JEGLIC is a Professor of Psychology at the John Jay College of Criminal Justice in New York, New York. Her research focuses on issues pertaining to the treatment and assessment of sex offenders and evidence-based public policy.