The reality about Internet-initiated sex crimes – those in which sex offenders meet juvenile victims online – is different, more complex, and serious but less archetypically frightening than the publicity about these crimes suggests.

• Mental health practitioners need an accurate assessment of the nature and prevalence of online child molestation because they are likely to encounter related issues in a variety of contexts.
  
• Psychologists who work with youth may come across some who have been victimized or are in danger because of off- or online sexual risk taking.
  
• Psychologists who work in settings such as schools may have to respond to the concerns of parents, teachers, and other adults.

In some cases, the publicity about Internet-initiated sex crimes may engender exaggerated fears that need to be dealt with by providing accurate information. In other cases, knowledgeable advice about how to handle situations in which youth are victimized, targeted, or at risk may be required. In addition, some psychologists may find themselves dealing with adults who are caught up in sexually compulsive online behavior that may include potential or actual illegal conduct with young people. Such offenders may come to treatment because they have been caught by law enforcement, but others may seek treatment voluntarily. For some, Internet-related problems may become apparent during individual, couple, or family therapy sessions.

The purpose of this article is to provide an accurate, research-based description of the characteristics and prevalence of this high profile social problem, make recommendations for effective responses, indicate needs for future research, and give professionals basic resources to help manage issues that arise in practice and other contexts

• (Table 1).

We present an overview of research relating to Internet-initiated sex crimes, much of it conducted by the authors at the Crimes against Children Research Center (CCRC) at the University of New Hampshire.

The CCRC research includes the first and second Youth Internet Safety Surveys (YISS-1 and YISS-2), telephone interviews with national samples of youth Internet users, ages 10 to 17, conducted in 2000 and 2005

• (Finkelhor, Mitchell & Wolak, 2000; Wolak, Mitchell & Finkelhor, 2006).

It also includes the National Juvenile Online Victimization (N-JOV) Study, the only research to date examining the characteristics of Internet-initiated sex crimes by interviewing law enforcement investigators

• (Wolak, Mitchell & Finkelhor, 2003a).

The N-JOV Study sheds light on the incidence and dynamics of Internet-initiated sex crimes in which online molesters were arrested by law enforcement and the characteristics of victims and offenders.

We surveyed a stratified random sample of 2,576 local, county, state, and federal law enforcement agencies between October 2001 and July 2002 to collect data about Internet-related sex crimes with juvenile victims

• (Wolak, Mitchell & Finkelhor, 2003b).

The goals were to

1. design a representative national sample of law enforcement agencies that would provide an overall picture of these crimes in the U.S.,
   
2. understand how these cases emerged and were handled in a diverse group of agencies, and
   
3. get detailed data about crime characteristics.
   

We used a two-phase methodology of mail surveys to agencies followed by telephone interviews. The mail surveys were sent to a stratified sample of agencies, with stratification based on specialization and training. Thus, the sample included high numbers of agencies that concentrated on these crimes or had specialized training, as well as a random selection of all other agencies.

The response rate to the mail survey was 88%, with 385 agencies reporting a total of 1,723 cases. We conducted telephone interviews with investigators about specific cases selected randomly from the cases reported in the mail surveys. The final data set, weighted to account for sampling procedures, includes data from 612 completed interviews, 129 of which concern Internet-initiated offenses.

• (Data collection for a 2nd N-JOV Study, replicating the first, began in June 2007.)

The research about Internet-initiated sex crimes makes it clear that the stereotype of the Internet child molester who uses trickery and violence to assault children is largely inaccurate

• (Wolak, Finkelhor & Mitchell, 2004).

Most Internet-initiated sex crimes involve adult men who use the Internet to meet and seduce underage adolescents into sexual encounters. The offenders use Internet communications such as instant messages, e-mail, and chatrooms to meet and develop intimate relationships with victims. In the great majority of cases, victims are aware they are conversing online with adults.

In the N-JOV Study, only 5% of offenders pretended to be teens when they met potential victims online

• (Wolak, et al., 2004).

Also, offenders rarely deceive victims about their sexual interests. Sex is usually broached online, and most victims who meet offenders face-to-face go to such meetings expecting to engage in sexual activity. Many victims profess love or close feelings for offenders.

In the N-JOV Study, 73% of victims who had face-to-face sexual encounters with offenders did so more than once. When deception does occur, it often involves promises of love and romance by offenders whose intentions are primarily sexual. Most offenders are charged with crimes such as statutory rape, that involve nonforcible sexual activity with victims who are too young to consent to sexual intercourse with adults

Media reports and Internet safety messages about Internet predators often suggest that online meetings between adults and youth which develop into sex crimes constitute a new dimension of child sexual abuse

• (e.g., Bahney, 2006; Filosa, 2007; Manalatos, 2007).

While a new medium for communication is involved, the nonforcible sex crimes that predominate as offenses against youth online are not particularly new or uncommon. All states have criminal laws deeming youth below certain ages too young to consent to intercourse

• (Glosser, Gardiner & Fishman, 2004; Manlove, Moore, Liechty, Ikramullah & Cottingham, 2005).

The crimes committed under these laws are referred to by a variety of names in different jurisdictions. For simplicity, we use the term statutory rape. Individual states vary considerably on ages of consent, with 16 as the most common demarcation, but ranging from 14 to 18

• (Norman-Eady, 2003).

Most states exempt peer relationships by requiring a minimum age for offenders or an age differential between the parties

• (Davis & Twombly, 2000).

Also, many states provide more severe charges or penalties for crimes involving younger youth, intoxicated youth, and adults in positions of trust with youth, such as teachers

• (Glosser, et al., 2004).

Violations of age of consent laws constitute a substantial proportion of sex crimes against minors in general. Analyses of crime report data suggest that 25% of the sex crimes committed against minors and reported to police involve statutory rape, numbering an estimated 15,700 reports across the U.S. in 2000

• (Troup-Leasure & Snyder, 2005).

This estimate is by no means a full measure of the number of such crimes. Many sex crimes against minors, perhaps the majority, are never reported to law enforcement

• (Finkelhor, Ormrod, Turner & Hamby, 2005).

Statutory rape is certainly more underreported than forcible rape because it often involves adolescents who may not view these incidents as crimes or themselves as victims

• (Berliner, 2002; Lanning, 2002).

The number of youth involved is suggested by data gathered from young people via the Youth Risk Behavior Survey, which found in one state that about

• 1% of girls ages 11-12 and
  
• 3.5% of girls ages 13-15

reported incidents of nonforcible sex with boys or men who were five or more years older

• (Leitenberg & Saltzman, 2000).

A national study conducted in 2002 found that

• 6% of girls had their first intercourse at age 14 or younger.
  
• Of those, 21% reported partners who were six or more years older
• (Abma, Martinez, Mosher & Dawson, 2004).

Although small percentages, these rates could represent large numbers of youth, and, if comparable to numbers in other states, are similar in magnitude to the number of girls who experience forcible sexual assaults during a given year

• (Finkelhor, et al., 2005; Hines & Finkelhor, 2007).

Statutory rape laws have engendered controversy because of concerns about criminalizing consensual sexual activity between peers, whether underage adolescents voluntarily engaging in sex should be considered victims, and whether such laws are fairly enforced

• (Cheit & Breslow, 2006; Colb, 2004; Miller, Miller, Kenney, & Clark, 1998).

In light of these controversies, it is important to keep in mind that crimes charged as statutory rape are diverse in their dynamics

• (Hines & Finkelhor, 2007; Troup-Leasure & Snyder, 2005).

The participation of underage youth, while generally deemed voluntary, is voluntary in varying degrees. Some victims are pressured to engage in sex, and some are intimidated

• (Cheit & Breslow, 2006; Darroch, Landry & Oslak, 1999).

Compared to adults and even youth in their late teens (ages 17 to 19), younger adolescents have little experience with intimate relationships or romance

• (Weinstein & Rosenhaft, 1991).

They often lack the ability to negotiate effectively with partners about sexual activity

• (Ponton & Judice, 2004).

Young adolescents with older partners have high rates of coerced intercourse

• (Manlove, et al., 2005).

In spite of controversies about statutory rape laws, the preponderance of public policy efforts in recent years has been to strengthen and enforce them

• (Donovan, 1997).

In part, these efforts have been prompted by research linking high teen pregnancy rates to youth who have sex with older partners

• (Darroch, et al., 1999; Donovan, 1997);

and by high profile news stories about teachers, priests, and other adults in positions of authority taking advantage of youth in their charge

• (Frawley-O’Dea, 2007; Shakeshaft, 2004).

While state laws remain inconsistent in terms of age of consent and age discrepancies that are considered criminal, crime data suggest that law enforcement activity is concentrated on more serious cases. The crimes which result in arrest typically involve younger adolescents and adult offenders

• (Troup-Leasure & Snyder, 2005).

Consistent with this broader picture, the Internet-initiated sex crimes pursued to arrest by law enforcement typically involve adult offenders who are ten or more years older than their underage victims

• (Wolak, et al., 2004).

How much are Internet-initiated sex crimes contributing to statutory rape?

There were an estimated 6,594 arrests nationwide for statutory rape in 2000

• (Troup-Leasure & Snyder, 2005).

During about the same time period (July 1, 2000, to June 30, 2001) federal, state, and local law enforcement agencies made an estimated 500 arrests for Internet-initiated sex crimes, 95% of which were nonforcible

• (Wolak, et al., 2003a; Wolak, et al., 2004).

If these Internet-initiated sex crimes were counted among the 6,594 arrests for statutory rape, which they may have been, Internet-initiated sex crimes would have accounted for approximately 7% of all statutory rapes.

This proportion of arrests may have grown since 2000 as Internet use has become more widespread, and more law enforcement agencies have been trained to respond to Internet- related crimes. In the context of general sex crime risk, however, these numbers suggest that Internet-initiated sex crimes account for a salient but small proportion of statutory rape offenses and a relatively low number of the sexual offenses committed against minors overall.

Many of the media stories and much of the Internet crime prevention information available suggest that it is naïve and inexperienced young children who are vulnerable to online child molesters

• (e.g., Blustein, 2007; Boss, 2007; Crimaldi, 2007; Manolatos, 2007).

However, 99% of victims of Internet-initiated sex crimes in the N-JOV Study were ages 13 to 17 (M = 14.46, SD = .14), and none were younger than 12

• (Wolak, et al., 2004).

Forty-eight percent were ages 13 or 14. This is a victim age profile that spans some important developmental shifts, but it is a considerably more restricted age profile than for conventional offline child molestation, which includes a large proportion of victims younger than 12

• (Finkelhor & Baron, 1986; Snyder, 2000).

Although adolescent immaturity may play an important role in the victimizations these youth experience, it is undoubtedly a different type of naiveté than that of preadolescent children. This distinction is important for developing effective prevention strategies.

First, the characterization of young people as vulnerable because of naïveté about the Internet itself is not accurate. By early adolescence (ages 12 to 13), youth Internet users generally understand the social complexities of the Internet at levels comparable to adults, when answering questions about good and bad things that can happen online and the need to exercise care

• (Yan, 2006).

Then, as youth get older and gain experience online, they engage in more complex and interactive Internet use

• (Livingstone, 2006).

This actually puts them at greater risk than younger, less experienced youth who use the Internet in simpler, less interactive ways.

Among youth ages 12- to 17-years-old, it was those ages 15 to 17 who were most prone to take risks involving privacy and contact with unknown people

• (Livingstone, Bober & Helsper, 2005).

Second, the characterization of young people as vulnerable because they are innocent about sex does not capture the nature of the sexual issues that get youth into trouble online. The reality of adolescent sexual development includes growing sexual curiosity, knowledge, and experience as youth make the transition from childhood to adulthood

• (Ponton & Judice, 2004; Steinberg & Morris, 2001).

Even in early adolescence, most youth are quite aware of, interested in, and beginning to experiment with sex

• (DeLameter & Friedrich, 2002; Weinstein & Rosenhaft, 1991).

By mid-adolescence, most have had romantic partners and are preoccupied with romantic concerns

• (Steinberg & Morris, 2001).

While most youth do not become sexually active in their early teen years, the median age for first intercourse is around 17

• (Guttmacher Institute, 2006).

Data gathered from teens in 2002 showed that

• 30% of girls and
  
• 32% of boys

had intercourse before age 17, with variation by race and gender

• (Abma, et al., 2004).

Among girls,

• 25% of those who were Hispanic,
  
• 30% who were non-Hispanic White, and
  
• 41% who were non-Hispanic Black

had intercourse before age 17.

Among boys, the numbers were

• 43% of those who were Hispanic,
  
• 25% of non-Hispanic White, and
  
• 53% of non-Hispanic Black.

Healthy romantic relationships and sexual development are the issues of concern when considering youthful vulnerability to online molesters. This is so for several reasons.

First, face-to-face peer relationships are the context in which most youth learn to handle the decisions, emotions, and negotiations of romance and intimacy

• (Connolly, Craig, Goldberg & Pepler, 2004).

The Internet-initiated sex crimes that are romances from the perspective of young victims typically take place in isolation and secrecy, outside of oversight by peers, family members and others in the face-to-face social networks of youth. This isolation may lead to relationships that form more quickly, involve greater self-disclosure, and develop with greater intensity than face-to-face relationships among peers

• (McKenna, Green & Gleason, 2002).

Second, a considerable portion of victims are in early and mid-adolescence. Few youth of those ages have the mature judgment and emotional self-regulation required to engage in healthy relationships that include sexual intimacy

• (Cauffman & Steinberg, 2000; Weinstein & Rosenhaft, 1991; Wolfe, Jaffe & Crooks, 2007).

Third, youth in their early and mid teens often struggle with emotional control

• (Mash & Wolfe, 2005).

When they are drawn into online relationships that include disclosures about sexual matters, the feelings that are generated may be particularly powerful and difficult to handle for youth just beginning to experience sexual desires.

Fourth, intense romantic and sexual involvements during early and mid-adolescence are associated with a range of negative outcomes (e.g., externalizing and risk behaviors)

• (Halpern, Kaestle & Hallfors, 2007; Neeman, Hubbard & Masten, 1995; Ponton & Judice, 2004)

and may result in neglect of other important developmental tasks, such as academic performance

• (Wolfe, et al., 2007).

Finally, early sexual activity is related to a variety of risk behaviors, both

• sexual (e.g., multiple partners, older partners, unprotected sex, early pregnancy) and
  
• otherwise (e.g., substance abuse, delinquency)

• (Ponton & Judice, 2004; Raj, Silverman & Amaro, 2000; Wolfe, et al., 2007).

These bode ill for youth in terms of mental health and academic achievement

• (Wolfe, et al., 2007).

In summary, what creates risk for teens online is not innocence about sex. The factors that make youth vulnerable to seduction by online molesters are complex and related to immaturity, inexperience and the impulsiveness with which some youth respond to and explore normal sexual urges.

Certain types of online interactions with unknown people make youth vulnerable.

The Internet is a dynamic, interactive environment that youth actively participate in creating

• (Greenfield & Yan, 2006),

and it is this aspect of the Internet that creates risks for youth who behave in specific ways. While many youth interact online with unknown people (i.e., people they do not know in person), most youth who do so are not at risk for sexual victimization

• (Wolak, Finkelhor & Mitchell, in press).

However, youth who send personal information (e.g., name, telephone number, pictures) to unknown people or talk online to such people about sex are more likely to receive aggressive sexual solicitations – those that involve actual or attempted offline contact

• (Mitchell, Wolak & Finkelhor, 2007b).

Aggressive solicitations do not necessarily involve sexual approaches from online molesters, and few youth who receive such solicitations agree to meet solicitors. Nonetheless, these findings from YISS-2 interviews with a nationally representative sample of youth Internet users are consistent with what we know about the dynamics of Internet-initiated sex crimes.

Online child molesters often seduce youth by using online communications to establish trust and confidence, introducing talk of sex, and then arranging to meet youth in person for sexual encounters

• (Wolak, et al., 2004).

Because of this, it makes sense that youth whose online interactions include sending personal information to and talking about sex with unknown people are more likely to encounter individuals who make online sexual advances and then try to move them offline.

Moreover, youth who send personal information and talk about sex are not typical youth Internet users. The majority of youth refrain from these behaviors

• (Wolak, et al., 2006; Ybarra, Mitchell, Finkelhor, & Wolak, 2007).

About three-quarters have not sent personal information online to people they do not know in person; only 5% talked online to unknown people about sex.

Visiting chatrooms is another interactive behavior that is related to receiving aggressive sexual solicitations

• (Mitchell, et al., 2007b).

Chatrooms allow for immediate, direct communications between participants, and many of those geared to adolescents are known for explicit sexual talk, sexual innuendo, and obscene language

• (Subrahmanyam,Smahel & Greenfield, 2006).

This atmosphere may attract online child molesters. Also, the youth who visit chatrooms may be more at risk than other youth.

There is some evidence that adolescents who visit chatrooms are more likely to

• have problems with their parents;
  
• suffer from sadness, loneliness or depression;
  
• have histories of sexual abuse; and engage in risky behavior than those who do not go to chatrooms
  
• (Beebe, Asche, Harrison & Quinlan, 2004; Sun, et al. 2005).

Youth who are lonely, shy, or lacking in social skills may interact with others in chatrooms to compensate for problems they have forming friendships offline

• (Peter, Valkenburg & Schouten, 2005).

Younger adolescents, in particular, may not be developmentally prepared to avoid or respond to the explicit sexual invitations they are likely to encounter in many chatrooms

• (Greenfield, 2004).

Most of the online child molesters described in the N-JOV Study met their victims in chatrooms. In a 2006 study, about one-third of youth who received online sexual solicitations had received them in chatrooms

• (Ybarra & Mitchell, in press).

Youth with histories of sexual or physical abuse, and other troubled youth, may be particularly vulnerable.

Youth Internet users with histories of offline sexual or physical abuse appear to be considerably more likely to receive online aggressive sexual solicitations

• (Mitchell, Finkelhor & Wolak, 2001; Mitchell, et al., 2007b).

Abused youth are more at risk for sexual victimization and exploitation in a variety of ways

• (Finkelhor, Ormrod & Turner, 2007; Raj, Silverman & Amaro, 2000).

Abuse history could be related to emotional needs or developmental distortions that make some youth less able to assess and more responsive to inappropriate sexual advances

• (Berliner & Elliott, 2002; Rogosch, Cicchetti, & Aber, 1995).

Some such youth may be vulnerable to online sexual advances because they are looking for attention and affection

• (Lanning, 2002).

In addition, childhood trauma is associated with adolescent risk behavior, including risky sexual behavior

• (Wolfe, et al., 2007).

Further, the youth interviewed for YISS-2 who engaged in high risk interactive behavior had high rates of a variety of offline problems, including

• rule-breaking behavior,
  
• depression, and
  
• social interaction problems at the clinical or borderline level as measured by the Child Behavior Checklist

• (Wolak, et al., in press).

So, the youth most at risk may exhibit a wide range from problems. For some, prior abuse may trigger risky sexual behavior that directly invites online sexual advances. But delinquency, depression, and social interaction problems unrelated to abuse also may increase vulnerability.

Posting personal information online does not, by itself, appear to be a particularly risky behavior.

Posting personal information online is widely regarded as putting youth at risk for victimization by online child molesters, but findings from YISS- 2 suggest it is not, by itself, associated with being sexually solicited online

• (Mitchell, Wolak & Finkelhor, in press; Ybarra, et al., 2007).

Despite admonitions against it, posting personal information is prevalent among youth Internet users. More than half of youth ages 10 through 17 who were interviewed in 2005 about behavior in the previous year had posted personal information online in the form of names, school names, ages, pictures of themselves, or telephone numbers

• (Wolak, et al., 2006).

In general, behaviors manifested by large numbers of people fail to predict events that are relatively uncommon. Consequently, it is not surprising that a common activity such as posting personal information online does not predict a relatively rare event such as receiving an aggressive sexual solicitation.

In addition, a survey conducted in 2007 found that 55% of youth Internet users ages 12 to 17 had profiles posted online

• (Lenhart & Madden, 2007).

Youth who created profiles or posted photos of themselves online were more likely to be contacted online by unknown people (of any age), but not more likely to get contacts they described as scary or uncomfortable

• (Lenhart & Madden, 2007; Smith, 2007).

So, while Internet safety advocates worry that posting personal information exposes youth to online molesters, we have not found empirical evidence that supports this concern. It is interactive behaviors, such as conversing online with unknown people about sex, that more clearly create risk.

Nonetheless, so far the data are quite general and caution should be used in interpreting a slim literature. There may be risks associated with posting particular kinds of information or posting in particular venues that research has not discerned.

For example, a 2006 content analysis of publicly viewable web pages posted by adolescents at the social networking site MySpace found that 5% included pictures of youth wearing swim suits or underwear

• (Hinduja & Patchin, in press).

Youth who post such sexually suggestive photographs may be more likely to receive online sexual solicitations. On the other hand, some of these youth may have other traits that explain their actions, such as histories of sexual abuse, which are associated with sexual risk-taking

• (Raj, et al., 2000).

In such cases, the Internet may be a mode of risk transmission rather than a creator of risk.

Social networking sites like MySpace do not appear to have increased the risk of victimization by online molesters.

Starting in early 2006, there was considerable publicity about the potential dangers of social networking sites

• (e.g., Apuzzo, 2006; Bahney, 2006),

which have become increasingly popular with adolescents.

By the end of 2006,

• 55% of youth Internet users ages 12 to 17 employed such sites, and
  
• 70% of girls ages 15 to 17 (Lenhart & Madden, 2007).

Fears among parents, child advocates, and law enforcement seem to have arisen particularly in regard to the amount of personal information about young people available at such sites. Media stories have suggested that online molesters could use the information youth post about their identities and activities to locate and stalk youth

• (e.g., Filosa, 2007; Roeper, 2006).

Nonetheless, a close perusal of media stories suggests that online molesters have not changed their tactics as a result of the advent of social networking sites

• (e.g., Rawe, 2006; Schrobsdorff, 2006).

In addition, between June and October 2007, we conducted over 400 interviews with police about Internet-related sex crimes in conjunction with a second N-JOV Study, and we have yet to find cases of sex offenders stalking and abducting minors based on information posted on social networking sites. Online molesters do not appear to be stalking unsuspecting victims, but rather continuing to seek youth who are susceptible to seduction.

Findings from YISS-2 suggest that maintaining online blogs or journals, which are similar to social networking sites in that they often include considerable amounts of personal information and pictures, is not related to receiving aggressive sexual solicitations unless youth also interact online with unknown people

• (Mitchell, et al., in press).

In addition, youth with profiles on social networking sites, even those who were actively trying to meet new people, were no more likely than other online youth to have uncomfortable or scary contacts with unknown people

• (Lenhart and Madden, 2007; Smith, 2007).

Further, an online survey of a representative sample of over 1,500 youth Internet users conducted in 2006 found that youth were more likely to receive online sexual solicitations via instant messages or in chatrooms than through social networking sites

• (Ybarra & Mitchell, in press).

Similar to admonitions against posting personal information, suggestions that social networking sites are more dangerous for youth than other types of Internet activities are not substantiated by the small amount of relevant research available.

Vulnerability appears to be more distinguished by interactive behavior than by online location or posting personal information, which is a relatively passive activity. However, while these conclusions suggest that fears about social networking sites have been over-stated, caution should be used in interpreting this small amount of research about a new phenomenon.

Patterns of risky online behavior make youth vulnerable.

While posting information online, by itself, may not be as risky as some fear, there is increased risk associated with a pattern of different kinds of potentially risky online behaviors that includes posting personal information

• (Ybarra, et al., 2007; Wolak, et al., in press).

The pattern was identified by looking at nine online behaviors that are often deemed, or could be, risky for youth to engage in (e.g.,

• interacting online with unknown people,
  
• having unknown people on a buddy list,
  
• talking online to unknown people about sex,
•  seeking pornography online,
  
• being rude or nasty online).

As the number of different types of these behaviors increased, so did the odds of online interpersonal victimization (i.e., sexual solicitation or harassment)

• (Ybarra, et al., 2007).

Youth who engaged in three or four different types of these online behavior were 5 and 11 times, respectively, more likely than those with none to report online interpersonal victimization.

Further, youth who interacted online with unknown people and also engaged in a high number of different risky online behaviors were much more likely to receive aggressive sexual solicitations than youth who interacted online with unknown people, but restrained their risky behaviors

(Wolak, et al., in press).

This profile of Internet victims as youth who take risks online is consistent with research from offline environments showing risk-taking youth to be more vulnerable to victimization

• (Jensen & Brownfield, 1986; Lauritsen et al., 1992),

Girls are more vulnerable, as are boys who are gay or questioning.

Girls are considerably more at risk than boys for victimization by Internet-initiated sex crimes, as well as for statutory rape in general

• (Cheit & Braslow, 2005; Troup-Leasure & Snyder, 2005; Wolak, et al., 2004).

Also, girls who become sexually active during early adolescence may be especially vulnerable because they are more likely to be involved with older partners

• (Leitenberg & Saltzman, 2000; Leitenberg & Saltzman, 2003; Manlove, et al., 2005)

and to engage in risky sexual behavior

• (Ponton & Judice, 2004).

While girls constitute a higher proportion of victims than boys, boys who identify as gay or are questioning their sexual orientation may be another population particularly susceptible to online victimization.

Boys constitute 25% of victims in Internet-initiated sex crimes, and virtually all of their offenders are male

• (Wolak, et al., 2004).

While being sexually victimized by male offenders does not confirm that male victims are gay, in the N-JOV Study, most of the Internet-initiated cases involving boys had elements that made it clear victims were gay or questioning their sexual orientation (e.g., meeting offenders in gay-oriented chatrooms).

Hostility and social stigma toward homosexuality

• (Tharinger & Wells, 2000; Williams, Connolly, Pepler & Craig, 2005)

as well as feelings of isolation and loneliness

• (Martin & D’Augelli, 2003; Sullivan, 2002)

may impair the ability of boys who identify as gay or questioning to form age-appropriate, intimate relationships. Concerns about confidentiality and feelings that problems are too personal to disclose can also limit their willingness to get information about sexual matters from trusted adults

• (Dubow, Lovko & Kausch, 1990).

For these reasons, some gay boys turn to the Internet to find answers to questions about sexuality or meet potential romantic partners, and there they may encounter adults who exploit them.

The widespread popularity of television shows such as “To Catch a Predator”

• (e.g., Hansen, 2006)

reveals the public fascination with online child molesters. The media has been quick to characterize such men as Internet or online “predators" and pedophiles

• (e.g., Minaya, 2006; Roeper, 2006).

Implicit in these characterizations is the notion that these are highly motivated and repetitive sex offenders who have deviant sexual interests in children and predilections to abduction and violent assault.

In fact, the considerable research and theory about child molesters – what impels them to offend, how likely they are to have large numbers of victims or re-offend, and whether they have violent propensities

• (Finkelhor, 1984; Knight, Carter & Prentky, 1989; Prentky, Janus & Seto; 2003; Ward, Polaschek & Beech, 2005)

– makes it clear that child molesters are, in reality, a diverse group that cannot be accurately characterized with one-dimensional labels. While there is little research specifically about online child molesters, there are indications that they occupy a narrow range on the spectrum of the sex offender population, one that largely excludes pedophiles and violent or sadistic offenders.

Online child molesters are generally not pedophiles.

Because online child molesters primarily target adolescents, not young children

• (Lanning, 2002; Wolak, et al., 2004),

such offenders do not fit the clinical profile of pedophiles who are, by definition, sexually attracted to prepubescent children (APA, 2000).

For several reasons, it would be difficult for pedophiles to use the Internet to target and recruit young children directly. Young children are not as accessible online as adolescents. They use the Internet less for communication, and they are more supervised in their online activities

• (Roberts, Foehr & Rideout, 2005; Wang, Bianchi & Raley, 2005).

Also, they are less likely to respond to overtures from online child molesters because they are, for developmental reasons, less interested in relationships, sex, and romance than adolescents

• (DeLamater & Friedrich, 2002).

While cases of pedophiles using the Internet to meet prepubescent victims directly are quite rare, such offenders do use the Internet in other ways. For example, some pedophiles get access to young child victims through online contact with parents or other adult offenders, or they use the Internet to acquire child pornography.

While most online child molesters do not appear to be motivated by pedophilia, some may have a sexual compulsion for or a primary sexual attraction to adolescent boys or girls. The term “ephebophile,” is sometimes used to refer to men attracted to adolescent boys, and “hebephilia” to denote attraction to adolescents of either sex, although these terms are used inconsistently

• (Nunez, 2003).

Furthermore, although sexual activity between adults and young adolescents is illegal, hebephilia is not a paraphilia according to the Diagnostic and Statistical Manual of Mental Disorders (1994); nor is it a deviant sexual attraction in the same sense as pedophilia, because adolescents are sexually mature. Rather it is a violation of legal and social norms

• (Berliner, 2002).

While there is little research about hebephilia, studies of adult men who seek adolescent girls in offline environments have found that they are more likely to have

• criminal histories,
  
• less education,
  
• feelings of inadequacy, and
  
• arrested psychosocial development

• (Hines & Finkelhor, 2007).

These offline offenders may be different from online child molesters, however.

Nunez (2003) suggests several possible motivations among adults who pursue sex with adolescents, which could apply to online molesters. They may

• seek admiration from victims who are sexually responsive but naïve,
  
• want to relive adolescent experiences,
  
• be inhibited by fear of adult partners, or
  
• desire the power and control they can exert over youth.

Some online child molesters may be primarily sexually attracted to adults but target adolescents for reasons that include

• impulse,
•  curiosity,
  
• anger, or
  
• desire for power
  
• (Lanning, 2001).

For some of these offenders, the danger and excitement of seducing underage youth is itself a source of sexual arousal

• (Carnes, 2001).

Online child molesters are rarely violent.

Violence is rare in Internet-initiated sex crimes. The evidence from the N-JOV Study

• (Wolak, et al., 2004)

suggests that online molesters are not among that minority of child offenders who abduct or assault victims because they have sadistic tendencies or lack the interpersonal skills to gain the confidence and acquiescence of victims

• (Lanning, 2002).

Most online child molesters are patient enough to develop relationships with victims and savvy enough to move those relationships offline

• (Wolak, et al., 2004).

They know what to say to teens to gain their trust, arouse their sexual interest, and maintain relationships through face-to-face meetings

• (Lanning, 2002).

Abduction is also rare. None of the victims in the N-JOV Study were abducted in the sense of being forced to accompany offenders

• (Wolak et al., 2004).

However, about one-quarter of the cases started with missing persons reports because victims ran away to be with offenders or lied to parents about their whereabouts. So, in many cases, abduction may have been feared.

This is not to say that online child molesters are never violent or never abduct.

In the N-JOV Study, 5% used threats or violence, mostly forcible or attempted rape

• (Wolak, et al., 2004).

In at least one highly publicized Internet-initiated case, a 13-year-old girl was murdered

• (CBS News, 2002).

Abductions have occurred as well, but are very unusual.

Overall, what we know about online child molesters suggests that they are not generally impulsive, aggressive, or violent.

One possible explanation is that Internet use in the early 2000s was concentrated among those with technical skills, higher educations, and higher incomes, statuses hard to attain by those with impulsive and violent inclinations. This pattern may be changing as Internet access and skills become more widely disseminated.

On the other hand, the Internet may never be conducive to antisocial offender styles involving impulse and intimidation, because initial interactions are remote, physical contact is not certain, and intimidation may be difficult to project. Nonetheless, video web cameras and other technologies that introduce sight and sound into online communications are becoming more sophisticated. It remains to be seen whether they could make the Internet a more attractive venue for antisocial offenders.

The offenders in undercover “sting” operations are somewhat different than those arrested for victimizing actual youth.

In the year covered by the N-JOV Study, more online molesters were arrested for soliciting undercover investigators posing online as adolescents, than were arrested for soliciting actual youth

• (Wolak, et al., 2003a).

These offenders also appeared to be different from other online molesters to some extent

• (Mitchell, Wolak & Finkelhor, 2005).

Those who solicited undercover investigators were somewhat older and more middleclass in income and employment compared to those who solicited actual youth. They were also somewhat less likely to have prior arrests for sexual offenses against minors or for non-sexual offenses, or to have histories of violence or deviant sexual behavior.

However, both groups had equally high rates of child pornography possession, about 40%, and rates of substance abuse, about 15%.

Moreover, one in eight offenders arrested in undercover operations had also committed crimes against actual youth victims, which were discovered as a result of the undercover operation. It may be that the offenders most likely to be fooled by undercover investigators lack suspicion about law enforcement because they have less criminal experience and higher social status. It could also be that some such individuals are less experienced or skilled and more naive in their pursuit of youth and are thus more easily caught.

Child pornography and exhibitionism are factors in Internet-initiated sex crimes.

Child pornography (i.e., sexually explicit images of children younger than 18) may play a role in Internet-initiated sex crimes that is different from the role it has played in offline sexual offending

• (Carnes, 2003; Jenkins, 2001; Taylor & Quayle, 2003).

In the N-JOV Study, 39% of online child molesters possessed child pornography

• (Wolak, Finkelhor & Mitchell, 2005a),

a felony under federal law and in most states.

While child pornography is not a new phenomenon, the advent of the Internet has changed its nature as a crime problem. Child pornography possession, which used to be seen as a low incidence crime committed almost exclusively by those with an enduring sexual interest in children

• (Lanning & Burgess, 1984),

has evolved into a more general crime problem with an increasingly diverse array of offenders who can access and circulate images easily and privately from home computers

• (Wolak, et al., 2005a).

The small amount of research about the motivations of child pornography possessors suggests that, among other purposes, it is used to fuel sexual fantasy, enhance masturbation, groom and seduce victims

• (Jenkins, 2001; Taylor & Quayle, 2003),

and sometimes accessed out of curiosity or for its shock value.

Child pornography production is also an aspect of Internet-initiated sex crimes. One in five online child molesters in the N-JOV Study took sexually suggestive or explicit photographs of victims, or convinced victims to take such photographs of themselves or friends

• (Wolak, Finkelhor & Mitchell, 2005b).

In YISS-2, 4% of youth Internet users were asked to take sexual pictures of themselves and send them to online solicitors

• (Wolak, et al., 2006; Mitchell, Finkelhor & Wolak, 2007a).

Many of these requests appeared to constitute production of child pornography under federal statutes.

In addition, if youth comply with such requests, solicitors and others can circulate the images widely online with no possibility that circulation can be curtailed. This is a situation some youth might not have the foresight to understand or appreciate.

The Internet also may be particularly attractive to offenders with exhibitionistic tendencies, who can use web cameras to transmit images of themselves online.

In the N-JOV Study, 18% of online child molesters sent photos of themselves in sexual poses to victims

• (Wolak, et al., 2004).

In YISS-2, 6% of youth who were sexually solicited received such pictures from solicitors

• (Wolak, et al., 2006).

Several technological developments could facilitate continuing increases in child pornography possession, distribution and production, as well as the use of the Internet by exhibitionists.

• The growing capacity of computers and removable media to store images;
  
• wireless technologies that create mobile access to the Internet via portable devices;
  
• widespread access to digital photography, including cell phones and other handheld devises that take and transmit photos; and
  
• the increasing use of video web cameras in chatrooms and during one-on-one communications such as through instant messages
  

Concern about the Internet has fostered speculation that it may increase the number of “hands on” sex offenders and the number of youth victimized, above and beyond the growth in offenses related to possession of child pornography. While there are plausible mechanisms by which this could happen, they remain speculations as yet unsupported by research findings.

Does viewing child pornography and participating in Internet sex sites trigger sex offending that would not have otherwise occurred?

Some researchers who have treated online child molesters have suggested several Internet mechanisms that may promote offending.

One is the way the Internet facilitates exposure to child pornography, which some believe may evoke or promote criminal sexual interests that were deeply buried or nonexistent prior to such exposure

• (Carnes, 2003; Russell & Purcell, 2006).

Another is the existence of and easy access to Internet groups that explicitly or implicitly endorse the legitimacy of sexual interests in underage youth. In theory, such endorsements could encourage potential offenders to act on sexual urges that they have previously resisted

• (Galbreath, Berlin & Sawyer, 2002; Quayle & Taylor, 2001).

Some have argued that websites where offenders trade child pornography motivate some to molest and produce new images for trade and to gain status among offenders

• (Taylor & Quayle, 2003).

Another possible catalyst to sexual offending may be the anonymity that the Internet appears to afford offenders, who can groom and seduce victims from their homes under the assumption they will not be observed. This anonymity, combined with the high degree of arousal that results from online sexual stimulation, could lower internal restraints that would normally inhibit acting on inappropriate sexual urges, or it could trigger impulsive behavior

• (Carnes, 2003; Cooper, Delmonico, Griffin-Shelley & Mathy, 2004; Galbreath, et al., 2002; Quayle & Taylor, 2003).

Chatrooms and other online venues that allow for quick, easy contact with youth (or law enforcement posing as youth) may facilitate acting on such impulses.

Some therapists have described cases in which the compulsive use of Internet pornography and other sexually oriented sites was used to relieve stress or depression

• (Carnes, 2001; Cooper, et al., 2004).

This sexually compulsive Internet use may be characterized by rapid escalations in the amount of time spent at online sexual activities and the expansion of sexual interests into areas individuals never previously had, including the seduction of minors.

These are plausible hypotheses, but it is not clear how applicable or generalized these mechanisms may be. Evidence in support of these mechanisms might include findings that Internet offenders have less prior sexual offending and deviance and more child pornography exposure than other child molesters.However, we know of no such evidence at the current time.

Furthermore, possible contrary hypotheses also need to be entertained. For example, among some groups of predisposed individuals, easy access to a wide variety of engrossing and high-quality child pornography could serve as a substitute for involvement with actual victims. A similar hypothesis was raised about pornography (but not specifically child pornography) prior to the advent of the Internet, but little evidence has been accumulated in its support

• (Kuchinsky & Snare, 1999).

Does the Internet make youth more accessible to child molesters?

The Internet may facilitate child molesting by making youth more accessible to offenders and creating opportunities for molesters to be alone with victims. It is possible that offenders can find youth more easily online than through conventional social networks. Online communications such as e-mail or instant messages allow frequent, swift, and private exchanges, which online molesters can use to develop relationships with and seduce victims outside of public view and parental supervision.

Moreover, youth may be more willing to talk extensively and about more intimate matters with adults online than in face-to-face environments. Disconcerting status differences such as age and social background that may pose barriers to comfortable face-to-face communications between adolescents and adults could be less of an obstacle online. The isolation of online interaction may engender feelings in both molesters and youthful victims that they are outside of inhibiting social constraints imposed by peers, parents, teachers, and others in their face-to-face social networks

• (McKenna, et al., 2002).

Feelings of intimacy may develop more quickly, and parties to online interactions may feel freer to broach sensitive or forbidden topics, like sex. Again, while such speculations are plausible, they need evidence before they can be considered valid assertions. For example, research might try to find out if youth victimized in Internet-initiated crimes are younger or have fewer or different risk factors than those victimized by similar crimes in offline environments.

Could the Internet have features that protect youth from victimization?

It is important not to exclude consideration of Internet features that might mitigate, as well as increase, youth vulnerability. One possible protective feature of the Internet is that time is required to move relationships from communication to physical contact. This allows for a period of reflection that may inhibit some dangerous or criminal acts that might occur in offline environments.

Another possible protective feature of the Internet is that it may be harder for adults online to project intimidation, authority, or other manipulative tactics that rely on physical presence and status differences. This may reduce the likelihood that adults can pressure young people into sexual activities online, in comparison to offline environments.

A third possibly protective feature is that Internet availability may have prompted adolescents to spend more time at home than they once did, because they can engage in a certain amount of adventure and risk-taking from a home-based computer. This may have reduced their exposure to sexual victimization in other, more risky environments, which they may have formerly frequented

• (Subrahmanyam, Greenfield & Tynes, 2004).

In addition, as Internet technology develops, some of the risk elements inherent in its use may decline. The introduction of video technology during online communications may inhibit some offenders, for example those who deceive victims about their ages or physical appearances. Also, some youth may be less vulnerable to online sexual advances if they have real-time visual images of offenders.

Sex crimes against youth have not increased.

An important fact that supports caution in speculating about how the Internet has facilitated child molestation is that several sex crime and abuse indicators have shown marked declines during the same period Internet use has been expanding.

From 1990 to 2005, the number of sex abuse cases substantiated by child protective authorities declined 51%, along with other related indicators

• (Finkelhor, in press; Finkelhor & Jones, 2006).

For example, the rate of sexual assaults reported by teenagers to the National Crime Victimization Survey declined by 52% between 1993 and 2005. A statewide survey of students in Minnesota also showed declines in sexual abuse during this period.

Other indicators that might reflect on sexual victimization have also improved. The rate of pregnancy among teenagers has declined; there have been fewer delinquency arrests, and fewer children running away from home

• (Finkelhor & Jones, 2006).

To claim, as one headline from Newsweek did, that the Internet has fostered a

• “shocking increase in the sexual exploitation of children,”

• (Nordland & Bartholet, 2001)

one has to explain why this epidemic has not been more apparent in aggregate indicators of juvenile sexual victimization.

One possibility is that sex offenders have migrated to the Internet from other environments, so that increases in online sex offending have been balanced by decreases in offline victimizations.

It is also possible that Internet-initiated sex crimes have increased dramatically but are still relatively few in number compared to offline sex crimes. If so, a serious rise in sex crimes facilitated by the Internet may come in the future as the Internet continues to expand its influence.

It may also be that the Internet is only affecting the subgroup of nonforcible sex crimes against adolescents, which are not well measured by most crime indicators because they typically focus on forcible offenses

• (Troup-Leasure & Snyder, 2005).

Or it could be that the Internet factors that hypothetically facilitate sex crimes are not as prevalent or powerful as some believe, or are counteracted by other factors that inhibit sex crimes.

Clearly, more research is needed on these issues. Because Internet-initiated sex crimes are a relatively new phenomenon, it may take some time before there is enough information to understand their role and relationship to juvenile sexual victimization overall. In the meantime, it is premature to talk about the Internet as an established facilitator of sex crimes, beyond the possession and distribution of child pornography.

Avoid descriptions of the problem that characterize victims as young children or emphasize violence and deception.

A considerable portion of media coverage of Internet dangers emphasizes young children as potential victims, or focuses on violence, abduction and deception. It may be more compelling in sex crime advocacy and prevention to typify victims as “innocent young children"

• (Best, 1990),

but this will not promote effective public policy or preventive behaviors among those most at risk. Portraying victims as preadolescents does not sensitize the public, parents, or adolescents to the nature of Internet-initiated sex crimes.

Similarly, characterizing Internet-initiated sex crimes as violent makes it hard for the public and professionals to recognize nonforcible crimes such as statutory rape, and it may prevent victims from reporting crimes which do not conform to violent stereotypes.

Further, it may lead practitioners to mistakenly assume force or deceit on the part of offenders, which can make victims reluctant to fully disclose what happened, and which can, in turn, compromise investigations and prosecutions.

It also may complicate the treatment of victims who cannot apply the violent stereotype to their situations. Although it would be a mistake to say that these crimes never involve violence or deception, the public already may be so aware of those possibilities that it is essential to provide countervailing information.

Be clear about why sex with underage adolescents is wrong.

Especially in a society where images of adolescent sexuality abound in the media, many adolescents and adults may not be clear that relationships between adults and underage adolescents are criminal.

It is valuable for the public to hear messages that reinforce norms and counteract media that present sexualized images of youth. This can include media stories that highlight the prosecution of statutory sex crimes, publicity about age of consent laws, and statements from authorities and opinion leaders about the rationale behind these prohibitions.

Important points to touch upon include

• the inequality of power and experience between youth and adults,
  
• the immaturity of teens and their lack of readiness for intimate relationships with adults, and
  
• the potential negative impact on victims in terms of healthy sexual development and other consequences.

Offenders and potential offenders need to hear a clear message that nonforcible sex with underage adolescents violates the social responsibility adults have toward youth for objective mentoring and custodianship.

Focus prevention efforts more on adolescents and less on parents.

Because parents are a receptive audience concerned about Internet safety, there is a strong inclination to direct prevention strategies toward them. But simply urging parents and guardians to control, watch, or educate their children may not be effective in many situations. The adolescents who tend to be the victims of Internet-initiated sex crimes may not themselves be very receptive to the advice and supervision of parents.

Moreover, some of the most vulnerable youth may be alienated from their parents, victims of familial abuse, or dealing with sensitive issues such as inner conflicts about sexual orientation that they feel their parents will not understand. Prevention strategies should be targeted more directly at adolescents themselves, using media and authorities, including other youth, that have their confidence.

As recent research on sexual assault has shown, another important audience for these messages may be peers and other "bystanders”

• (Banyard, Plante & Moynihan, 2004).

For example, peers who know about the romantic involvements of their friends may heed messages to take preventive measures when they see friends drawn into dubious online relationships. Participants in Internet networks and chatrooms are also bystanders who can take steps to report inappropriate behavior and enforce responsible standards.

Focus prevention frankly on concerns relevant to adolescents, including autonomy, romance and sex.

Prevention approaches need to acknowledge the independence and developmental interests of adolescents and to acknowledge that normal adolescent sexual feelings, urges, and curiosity are important factors in these cases.

Too often approaches to prevention shy away from realistic discussions about ordinary sexual feelings. Instead they advise things like,

• “Tell a parent, teacher, or trusted adult if you feel uncomfortable about anything you see on the Internet” (Internet Keep Safe Coalition, 2007).

Such characterizations gloss over the reality of adolescent sexual development. We recommend educating youth frankly about the dynamics of Internet-initiated and other nonforcible sex crimes. Youth need candid, direct discussions about seduction and how some adults deliberately evoke and then exploit the compelling feelings that sexual arousal can induce.

Even young adolescents should be given basic information about the inappropriateness of romantic advances from adults. This information should include reassurances that it is normal to have strong sexual feelings, but wrong for adults to provoke or exploit these feelings, especially with youth who are inexperienced in coping with sexual desire and intimate relationships.

Prevention should be developmentally appropriate and an aspect of broader programs that focus on healthy sexual development and avoiding victimization.

Ideally, prevention information about Internet-initiated sex crimes should be part of broader programs that teach youth about healthy sexual development, including how to recognize and avoid sexual victimization. Such programs should start in early adolescence and be developmentally geared.

With younger adolescents, programs might emphasize

• types of Internet use and web sites,
  
• risky situations youth may encounter online, and
  
• practicing refusal and resistance techniques.

Older adolescents could use information about

• age of consent laws,
  
• the problems of relationships with older partners,
  
• the dangers of transmitting sexual pictures, and
  
• online grooming tactics used by sexual offenders.

All such prevention messages need to be developed and tested in conjunction with youth themselves, especially because few if any programs of this sort currently exist and wide gaps may be present between how adults and young people view and understand the online environment

Focus prevention more on interactive aspects of Internet use and less on posting personal information.

There is no empirical evidence that posting personal information, by itself and independent of engaging in a pattern of online risky behavior, puts youth at risk for sexual victimization.

Further, millions of youth use social networking sites safely, and we have not found evidence that these sites are more risky than other online venues popular with youth.

Rather than focusing on types of online sites or non-interactive pursuits such as posting information, prevention messages should focus on online interactions because Internet-initiated sex crimes come about through direct communications between offenders and victims.

This includes educating youth about the specific kinds of Internet interactions that are most associated with victimization, such as talking online about sex to unknown people. At the same time, judicious online contact with unknown people is not harmful or dangerous

• (Wolak, Mitchell & Finkelhor, 2002; Wolak, et al., in press).

Overly broad admonishments about talking to strangers may be seen as unrealistic and undercut credibility. Instead, we should encourage youth to be wary about talking online with unknown people about sex and to report inappropriate sexual overtures to web site and law enforcement authorities.

Educate youth about criminal behavior and child pornography.

We also recommend giving adolescents specific, age appropriate information about the potentially criminal nature of many aggressive sexual solicitations. An adult using the Internet to make sexual advances to minors is a crime in most U.S. jurisdictions.

Prevention messages also need to publicize the prevalence and risks of adults asking youth to take sexually explicit pictures of themselves. We need to extend this education to parents and clinicians as well, so they understand the unsuspected and risky uses web and digital cameras may be put to and the potential magnitude of this problem.

Clinicians who work with youth who have been sexually abused and assaulted should be alert to the possibilities that illegal sexual images of victims may have been produced. Such images could be available on the Internet, and their possible wide-spread distribution poses ethical and therapeutic issues that extend beyond those considered in the typical sexual assault response and treatment paradigm

• (Holland, 2005; Palmer, 2005).

Develop targeted prevention approaches for the most at risk youth populations.

Research suggests that youth with histories of offline physical or sexual victimization are more likely to be victimized online. Youth who are alienated from and in conflict with their parents may also be vulnerable, as well as those who suffer from depression, have delinquent tendencies or struggle with social interaction problems. Boys who are gay or questioning their sexual orientation are another group that may be at high risk, as well as youth engaged in sexual risk taking.

Unfortunately, most current prevention programs are developed for the general population of youth Internet users, but the most at risk youth may be hard to reach with these generic materials. Since mental health professionals may have more opportunities for interaction with at risk youth, we suggest that mental health organizations, with youth input, work to develop prevention materials that can be used by practitioners with such populations, individually or in small groups.

Assess for a pattern of risky online behavior.

Our research has found that, while many individual Internet behaviors are not associated with higher rates of receiving aggressive sexual solicitations, there is increased risk associated with engaging in a pattern of online risk-taking. The more risk behaviors a youth engages in, the higher the risk.

The nine items, listed below in order of prevalence

• (Ybarra, et al., 2007),

suggest behaviors that may indicate a risk-prone pattern of Internet use. There are certainly others not identified here. However, this list covers the more common and recognizable risk behaviors, and these were the items which, when aggregated, were associated with higher risk in our research

• (Ybarra, et al., 2007).

This is not a developed and tested instrument, but the list can be used to quickly identify vulnerable youth who are engaging in an excess of risky Internet behavior. The least prevalent behaviors may pose the most concern.

•Posting personal information online, 56% of youth Internet users

•Interacting online with unknown people, 43%

•Having unknown people on a buddy list, 35%

•Using the Internet to make rude and nasty comments to others, 28%

•Sending personal information to unknown people met online, 26%

•Downloading images from file-sharing programs, 15%

•Visiting x-rated sites on purpose, 13%

In addition, forms of family intervention – focusing on family conflict, parent-child relationships, supervision – that have proven effective with delinquent youth may be useful with some parts of this population

• (Alexander, et al., 1998; Henggler, et al., 1998).

Connecting youth with positive sources of mentoring and education around sexual and relationship issues may also be helpful. Some practitioners may also have important roles in gathering forensic evidence from victims and obtaining their cooperation with law enforcement. Such practitioners need to develop skills for building rapport with adolescents and explaining the function of law enforcement in protecting other youth

• (Finnegan, 2006).

Treating offenders.

Somewhat more literature exists about treatment approaches with Internet sex offenders, growing out of experience in treating a variety of Internet-related behavioral problems

• (Carnes, 2003; Cooper, 2002; Quayle, Erooga, Wright,Taylor & Harbison, 2005; Quayle, Vaughn & Taylor, 2006).

As with all sex offenders, a key challenge is to recognize their diversity and make assessments that allow for treatment and social control options corresponding to the needs and problems of the offenders, while protecting potential victims.

There are likely to be individuals with serious sexual deviations, sex crime, and general crime histories in this population, as well as those with few mental health problems and no criminal history

• (Galbreath et al., 2002).

Efficacy in general sex offender populations has been demonstrated for a variety of child sexual abuser treatment programs based on cognitive behavioral and relapse prevention approaches. These programs target such problems as

• using sex to regulate emotional states,
  
• rationalizations and minimizations about victims, and
  
• social skill deficits that inhibit consensual relationships with peers
  (Yates, 2005).

Much of this is likely to be applicable to online offenders. If anything, the higher social functioning that may be required for recruiting and grooming victims via the Internet may make therapeutic approaches more successful with Internet offenders than with the general sex offender population.

Research needs

Very little is currently available in the way of research on which to base policy and practice in this field. Research takes on a particular value here because the online world changes quickly, and much of its activity is obscured from outside observation.

We suggest research be undertaken in four key directions.

One research direction should be organized to monitor the nature and dynamics of online sexual behavior involving youth. Media accounts sometimes alert the public and professionals to developing problems -- like juveniles solicited to produce online child pornography

• (Eichenwald, 2005)

-- but the online environment changes so quickly that formal methods of monitoring the extent and dynamics of new developments are badly needed.

For example, what will be the impact of new technologies, especially cell phones and other wireless handheld devices, on the dynamics and incidence of Internet-initiated sex crimes?

Research strategies should include ongoing surveys of nationally representative samples of youth and focus groups with young people about Internet activities. Online techniques may prove particularly appropriate for recruiting and surveying youth for such studies

• (e.g., Ybarra, et al., in press).

The research population should include the full range of children and youth who use the Internet independently. Longitudinal studies that capture developmentally based changes in Internet use are also needed.

Another research direction should focus on learning more about the development of Internet sex offending behavior. It will be useful to know if there are distinct new groups of offenders in this population, and whether the Internet has introduced new pathways into offending behavior.

Such studies may be helpful in developing strategies for preventing and short-circuiting Internet offending. Given the increasing law enforcement activity in this arena, researchers may be able to recruit pools of online offenders from prisons and treatment programs.

Soliciting practitioners who treat offenders to enter data into centralized data bases is a possible quantitative method which could allow for comparisons between off- and online offenders. Semi-structured interviews and similar qualitative methods would also provide helpful insights.

In addition, general population studies using both qualitative and quantitative methods that look at online sexual behavior would provide needed information about the online sexual compulsivity that appears to play a role in at least some Internet offending.

The third research direction should center on learning more about victims and potential victims. It will be useful to know about their characteristics, prior experiences and attitudes.

• How do potential victims view the Internet and the norms and standards they see as applying to themselves and other people in Internet interactions?
  
• Whom might they view as authoritative sources of information and advice about Internet activity?
  
• What needs do they have in the aftermath of disclosure and investigation?

Research has already suggested some of the risk-taking activities that may characterize potential victims

• (Ybarra, et al., 2007).

Surveys and focus groups with such risk-taking youth may provide insight into their behavior and suggest prevention approaches. Such research requires sensitivity. Recruiting and surveying victims through victim advocacy programs and children’s advocacy centers is one possible approach.

Finally, program and policy evaluation research should be undertaken.

• What are the impacts of different types of prevention messages and strategies?
  
• Are youth taking prevention messages to heart?
  
• Are they being implemented?
  
• Do specific technological strategies make any difference?

A wide range of Internet safety programs are being developed and disseminated, most with little evaluation. Experimental designs, with youth randomly assigned to programs, and outcome-based evaluations could easily be implemented to assess the value of such programs, which are often administered in schools and organized youth groups. Since online research techniques have greatly reduced the time and costs involved in such studies, there is little excuse for programs not to build in and report evaluation results as an integral part of their efforts. This process is crucial to avoid the potentially wasteful dissemination of programs that are not effective.

The research about Internet-initiated sex crimes indicates that the stereotype of the Internet ”predator” who uses trickery and violence to assault children is largely inaccurate

• (Wolak, et al., 2004).

Most Internet-initiated sex crimes involve adult men who use the Internet to meet and seduce young adolescents into sexual encounters. Most such offenders are charged with crimes, such as statutory rape, that involve nonforcible sexual activity with victims who are too young to consent to sexual intercourse with adults. Numbers suggest that Internet-initiated sex crimes account for a salient but small proportion of all statutory rape offenses and a relatively low number of the sexual offenses committed against minors overall.

While online molesters take advantage of developmentally normal adolescent interests in romance and sex, some youth may be particularly at risk. This group includes

• boys who are gay or questioning their sexual orientation;
  
• youth with histories of sexual or physical abuse; and
  
• those who frequent chatrooms, talk online to unknown people about sex, or engage in patterns of risky off- or online behavior.

While there is little research about online child molesters, they appear to occupy a restricted range on the spectrum of the sex offender population and include few true pedophiles or violent or sadistic offenders. We need frank, accurate prevention programs for youth, thoughtful treatment for victims, and continued research. As access to interactive Internet technologies broadens with the introduction and spread of wireless and handheld technologies, such as cell phones and personal organizers, youth Internet use could become harder to monitor, and accurate descriptions of and education about risks to youth will become even more important.

• Janis Wolak, Crimes against Children Research Center and Family Research Laboratory, University of New Hampshire;
  
• David Finkelhor, Crimes against Children Research Center and Family Research Laboratory, University of New Hampshire;
  
• Kimberly J. Mitchell, Crimes against Children Research Center and Family Research Laboratory, University of New Hampshire;
  
• Michele L. Ybarra, Internet Solutions for Kids, Inc, Santa Ana, CA.

This manuscript is largely based on three research projects conducted by the authors at the Crimes against Children Research Center, University of New Hampshire.

The first Youth Internet Safety Survey was funded by the U.S. Congress through a grant to the National Center for Missing & Exploited Children.

The National Juvenile Online Victimization Study and the second Youth Internet Safety Survey were funded by the National Center for Missing & Exploited Children and the U.S. Department of Justice, Office of Juvenile Justice and Delinquency Prevention.

Program support for the preparation of this manuscript was provided by the Verizon Foundation.

Points of view or opinions in this document are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice or other entities that provided support.

Thanks to the members of the Family Violence Research Seminar of the Family Research Laboratory and Crimes against Children Research Center at the University of New Hampshire for their valuable comments on an earlier version of this manuscript.

Correspondence should be addressed to Janis Wolak, Crimes against Children Research Center, University of New Hampshire, 10 West Edge Drive, Durham, NH 03824. E-mail can be sent to Janis.Wolak@unh.edu.