INTRODUCTION
There has been a boom in the development and use of different Information and Communication Technologies (ICTs) that are mediated by the Internet. ICTs, therefore, play an everincreasing role in the lives of adolescents, especially given that young people use them in multiple domains - to socialize, study, and play ( Livingstone & Smith, 2014). For example, in Chile --where this research was carried out-- 94% of homes with schoolchildren have Internet access ( SUBTEL & Brújula, 2017), although there are still differences between rural areas with less connectivity compared to urban areas ( Cadem, 2023) ICTs benefit society but may also give rise to risky behavior patterns, given that many criminal practices have transferred from the offline world to the online context ( Mayer, 2011). Among all online criminal practices, the study of online sexual abuse (OSA) against children and adolescents has been the most studied, given its high prevalence and its negative effects ( Toro, 2021).
OSA corresponds to all kinds of technology-mediated sexual abuse ( Quayle & Sinclair, 2012). These forms of OSA go from unwanted exposure to pornography to online grooming and online sexual exploitation ( Ramiro et al., 2019).
OSA can affect people of all ages, but children and adolescents have been the groups more attended by researchers because they correspond to one of the most vulnerable groups in society and tend to be the segment most pursued by online sexual offenders ( Livingstone & Smith, 2014). It is essential to find out more about dynamics within the adolescent age group, given that they use the Internet in riskier ways (e.g., impulsive use of the Internet, sharing personal information with strangers) ( Guerra et al., 2019a) and, in turn, have less parental supervision than younger children ( Marret & Choo, 2018).
In the United States, it is estimated that 9% of adolescents have received sexual solicitations from adults, 11% have experienced online harassment, and 23% have been exposed to pornography ( Jones et al., 2012). In Spain, it is estimated that 13% of adolescents have received sexual requests from an adult through the Internet, and 8% had some form of sexual interaction with an adult online ( De Santisteban & Gámez-Guadix, 2018). In Chile, it is estimated that 51% of adolescents have been exposed to unwanted pornography, 18% have been pressured to send sexualized personal images or videos ( Guerra et al., 2021), 15% have been victims of online sexual harassment, and 11% have received unwanted sexual proposals over the Internet (Pinto & Venegas, 2015).
The consequences of OSA are associated with severe symptoms and social adaptation problems (Guerra et al., 2020), like depression, anxiety, post-traumatic stress disorder (Bates, 2017; Döring, 2014), antisocial behaviors or substance abuse ( Houck et al., 2014), and greater suicidal risk ( Medrano et al., 2018).
Factors associated with online sexual abuse
The risk of OSA is higher in adolescents who present a psychologically more vulnerable profile, with poor offline social networks and a limited sense of well-being, since they tend to seek support on online networks, becoming easy prey for online aggressors ( Webster et al., 2012). These adolescents would tend to present a more problematic use of the Internet, including addictive patterns of use that would lead them to present more risky behaviors online, such as maintaining close contact with strangers, publishing personal data, or uploading compromising images that online offenders could later use engage them in online sexual activities ( Guerra et al., 2019b).
Additionally, adolescents with a low level of satisfaction with their lives, with family difficulties, with a lower perception of self-esteem and self-efficacy would be more likely to give personal information to strangers and thus be more exposed to the actions of online offenders (Alju‐boori et al., 2021; Jonsson et al., 2019). Some studies have also shown that the adolescents most vulnerable to abuse are those with the slightest knowledge of the different forms of OSA and how it differs from healthier ways of interacting online (Katz et al., 2018; Manrai et al., 2021).
For this reason, efforts have been made to prevent OSA, including considering the need to reduce time spent using the Internet, avoiding risky behaviors, balancing the greater well-being associated with online activities with of‐fline activities (e.g., face-to-face social interactions), as well as promoting knowledge of Internet risks and coping strategies for potential situations of OSA, the latter as a way of providing adolescents with greater personal resources that promote their self-efficacy and provide them with skills to prevent and disclose OSA ( Aljuboori et al., 2021).
1.2 Prevention strategies
The prevention of OSA has been approached from different angles. There are laws and programs to dissuade potential sex offenders from committing OSA ( Hudson, 2017), as well as technological methods --software to block cyber-attacks-- ( NetClean, 2018). There are also recommendations for children regarding ways to avoid exposing themselves to online risks and to do an early disclosure in case they suffer from OSA ( Patterson et al., 2022). Among them, the work that UNICEF has been carrying out stands out, who have published a series of awareness guides on "digital coexistence" in which they provide suggestions to children and adolescents for the responsible use of the Internet ( UNICEF, 2016, 2020).
The role of caregivers has also been highlighted. Parents should be competent in the knowledge of the online world so that they can supervise and control what children and adolescents do on the Internet ( Livingstone & Smith, 2014; Whittle et al., 2013). However, excessive control and supervision can cause children not to trust their parents for fear of reprisals and continue exposing themselves to risks behind their parents' backs ( Manrai et al., 2021).
Consistent with the above, Jonsson et al. (2019) compared the parental relationships of Swedish adolescent victims of OSA with those adolescents who were not victims and reported that victims had poorer relationships with both parents, experienced less parental care, and more overprotection. Fleming et al. (2006) found that control measures, such as the use of blocking software without adequate parental communication, are ineffective. This underpins why there is a tendency to recommend that parents balance control with support and acceptance to supervise their children and allow young people to feel understood and loved ( Toro & Guerra, 2021).
Research suggests that parents should guide children to increase their knowledge about OSA in order to be able to identify risky situations since young people have difficulties in recognizing such situations because aggressors often use strategies of gradual approach that prevent victims from recognizing the danger until it is too late ( Manrai et al., 2021). Additionally, it is necessary to support adolescents to be able to identify OSA and know how to prevent and react to it, increasing their sense of self-efficacy to face this type of situation, such as it occurs in offline sexual abuse ( Guerra et al., 2018).
1.3. The present research
Although the problem of OSA is global, there is an interest in having local information from the Chilean context to generate prevention measures that adjust to local needs. For this reason, the objectives of this research are: i. To evaluate the relation between risk (frequency of Internet use) and protective factors against OSA (knowledge about OSA, well-being, and self-efficacy) in adolescents; ii. To evaluate the relation between parental characteristics (supervision, support, and Internet skills) and the already mentioned risk and protective factors; and iii. Find out if adolescents with parents with different profiles (based on different combinations of parental supervision, support, and Internet fluency) present differences in risk and protective factors against OSA.
We hypothesize i. that the risk factor will be inversely associated with the protective factors; ii. Parental supervision will be inversely associated with the risk factor; iii. Parental support and Internet skills will be directly associated with protective factors; iv. Adolescents with parents with more balanced profiles will have lower risk factors and higher protective factors.
METHODS
2.1.Participants
Participants were 82 high school students (12-17 years old; M= 14.00; SD= 1.05) from Melipilla, Chile (65.9% female, 13.4% nonbinary gender, 12.2% prefer not to say their gender, and 8.5% male).
2.2. Instruments
We used two instruments to evaluate adolescents' perception of parental characteristics and four to evaluate adolescents' risks/protective factors.
2.2.1. Perception about parent's competence in Internet and parental supervisión
This instrument included two independent questions. Adolescents were asked to rate their parents' Internet skills (0= they do not know anything, to 4= they know how to use the Internet very well) and to what extent their parents supervise their children's Internet use (0= not at all, to 4= always).
2.2.2. The Family Support subscale of the perceived social support scale (Zimet, 1988)
This instrument has four items that evaluate adolescents' perception of parental support (1 = rarely to 4 = always or almost always). The total score is obtained from the sum of the four items (possible values between 4 and 16; higher punctuation refers to a higher perception of support). In this study, the reliability of the subscale was adequate (Alpha= 0.81).
2.2.3. Open-ended questionnaire on the knowledge of adolescents about OSA.
There are three open questions to evaluate the knowledge that adolescents have regarding OSA (What is OSA? What are the characteristics of online sexual offenders? What do you think are the characteristics of victims of OSA?). The answers were scored by two independent evaluators as follows: 3 points when the adolescent provides a detailed explanation; 2 points when they give a correct but very partial explanation; 1 point when they mention related concepts but leave out distinctive aspects of the phenomenon; 0 points when they give a wrong answer. The total score is obtained from the sum of the points of the three questions (possible total values between 0 and 9). Interrater reliability was appropriate (Krippen‐dorff's Alpha= 0,72).
2.2.4. Internetrelated experiences questionnaire (Bera‐nuy et al., 2009).
It is a 10-item scale that evaluates excessive use of the Internet. The answers are on a scale ranging from 1= not at all to 4= a lot. The possible scores vary between 10-40 points; higher scores indicate addictive Internet use (Alpha= 0.73).
2.2.5. The scale of satisfaction with life (Atienza et al., 2000)
This instrument consists of five items that evaluate subjective well-being on a scale from 1 to 5, with 1 being strongly disagree and 5 being strongly agree. The scores of the responses are added to obtain the values (values between 5 and 25; higher values mean higher perception of well-being) (Alpha= 0.81).
2.2.6. Self-efficacy scale (Schwarzer & Jerusalem, 1995).
This consists of 10 items that evaluate perceptions of self-efficacy. Each item is in a Likerttype response format, with 1= almost never and 4= always or almost always. The responses must be added to obtain the total values (values between 10 and 40; higher values mean higher self-efficacy) (Alpha= 0.90).
2.3. ProcedureThe ethics committee of the Universidad Santo Tomás approved the project.
The participation of an educational establishment in the Metropolitan region was requested, and coordination was carried out with the director and social worker. Informed consent was sent to parents (or guardians) and adolescents, including the link to access the instrument. Participation was requested from all secondary education students (N= 290). Of these, 82 students answered the instruments anonymously through an online platform.
2.4. Data análisis
We performed descriptive and correlation analysis with the study variables and then a hierarchical cluster analysis (squared Euclidean distance and the Ward method) to identify groups of parents with different profiles (based on child perception of parental support, supervision, and Internet competence). Finally, Anovas was performed to explore the differences between the resulting groups of parents on OSA risk (excessive use of the Internet) and protective factors (self-efficacy, well-being, and knowledge regarding OSA). All the analyses were performed in SPSS.
RESULTS
3.1. Adolescents' variables
On average, adolescents score 22.98 (SD=5.14) on the frequency of Internet use scale. Considering the possible values of the scale (from 10 to 40 points), it is estimated that Internet usage is moderate. In the same way, when considering the possible values of the self-efficacy scale (10 to 40 points), it is estimated that the participants present a moderate to high Self-efficacy (M=28.03; SD=6.77). Doing the same exercise, it is estimated that the participants present moderate to high well-being (M=17.79; SD=4.38, possible values between 5 and 25) but a low knowledge about OSA (M=3; SD= 1.48, possible values between 0 and 9).
The correlations show that the risk factor (frequency of Internet use) is inversely associated with two of the three protective factors (self-efficacy and well-being) but not with adolescents' knowledge about OSA. This indicates that adolescents with lower self-efficacy and well-being present more problematic Internet use. See Table. 1
3.2. Parent's variables
On average, adolescents perceive their parents as moderately supportive (M=11; SD=3.04; possible values from 4 to 16), with moderately high Internet supervision (M=2.76; SD=1.01; possible values from 0 to 4), and with moderate to high Internet skills (M= 2.83; SD=1.07).
It is appreciated that parental support is directly associated with the three protective factors: self-efficacy, well-being, and adolescent's knowledge about OSA, which means that the adolescents with the most protective factors against OSA are the ones with more supportive parents. In turn, parental supervision is inversely related to the frequency of Internet use by adolescents and directly to their sense of self-efficacy and well-being.
The cluster analysis retained three clusters of patenting styles, significantly different in parental support (F(79,2)= 186.223; p.01) and parental supervision (F(79,2)= 20.576; p.01) but with a similar level of Internet parental skills (F(79,2)= .438; p=.65). The first cluster was named as balanced parents since it grouped 37 adolescents with high parental support and high parental supervision. The second cluster was called controlling parents since it included 33 adolescents with parents with high supervisión but with medium to low parental support. The third cluster was called negligent parents and grouped 12 adolescents with low parental support and low parental supervision. See Figure 1. but with medium to low parental support. The third cluster was called negligent parents and grouped 12 adolescents with low parental support and low parental supervision. See Figure 1.
| Min-Max | Mean (SD) | 2. FS | 3. PS | 4. IPS | 5. Use I. | 6. Self-E | 7. Wellb. | |
|---|---|---|---|---|---|---|---|---|
| Parents/Family | ||||||||
| 1.Family Support | 5-16 | 11(3.04) | - | - | - | - | - | - |
| 2.Parental Supervision | 0-4 | 2.76(1.01) | .52** | - | - | - | - | - |
| 3. Internet parental skills | 1-4 | 2.83(1.07) | .09 | .23* | - | - | - | - |
| Adolescent | ||||||||
| 4. Internet use | 11-38 | 22.98(5.14 | -.10 | -.23* | -.11 | - | - | - |
| 5. Self-efficacy | 13-40 | 28.03(6.77) | .42** | .27* | .07 | -.34** | - | - |
| 6. Wellbeing | 5-25 | 17.79(4.38) | .51** | .23* | -.02 | -.24* | .49** | - |
| 7. Knowledge about OSA | 0-7 | 3(1.48) | .23* | .10 | .07 | -.19 | .03 | .10 |
Notas: rho \ p.05, **p.01, SD= Standard deviation*
| Balanced | Controlling | Negligent | F(79,2) | |
|---|---|---|---|---|
| Use of Internet | 22.19 | 23.30 | 24.50 | 1.026 |
| Self-efficacy | 30.81 | 25.42 | 26.67 | 6.618** |
| Wellbeing | 19.65 | 16.09 | 16.75 | 7.078** |
| OSA-Knowledge | 3.51 | 2.73 | 2.17 | 5.154** |
Notas: **p .01
There were no differences in the frequency of Internet use among adolescents in the three groups. However, adolescents from balanced parents have the highest scores for self-efficacy, well-being, and OSA knowledge. Post hoc analyses (Scheffe test) show that cluster 1 (balanced) scores significantly higher in self-efficacy and well-being than cluster 2 (controlling), but these differences are not significant compared to cluster 3 (negligent). Additionally, there are significant differences in Internet knowledge between clusters 1 and 3 but not between clusters 1 and 2. See Table 2.
DISCUSSION
This research wanted to evaluate the relationship between adolescents risk and protective factors against OSA abuse, and between those and specific parental characteristics. The results suggest that adolescents with higher self-efficacy and sense of well-being are those who present less problematic use of the Internet, which is consistent with previous studies ( Aljuboori et al., 2021; Guerra, 2019; Webster et al., 2012) and allows us to approve hypothesis 1 of this study.
Regarding the parental variables, the results indicate that, on average, adolescents perceive their parents as moderately supportive, with moderately high Internet supervision and moderate to high Internet skills. It is striking that adolescents perceive their parents good skills on the Internet, unlike previous reports, which alluded to the digital gap between adolescents and parents ( García et al., 2007). The results of this study call for further investigation into whether the digital gap is still a problem or could already be considered overcome, understanding that parents do have the necessary skills to support and supervise their children on the Internet.
Correlation analysis supports hypothesis two since parental supervision was inversely related to the frequency of Internet use by adolescents. This confirms what other authors have said regarding the need for parents to be aware of and monitor what their children do on the Internet, as this would reduce risk factors ( Livingstone Smith, 2014; Whittle et al., 2013).
Hypothesis three received partial support since only parental support was directly related to the three protective factors (self-efficacy, well-being, and adolescents knowledge about OSA), but Internet parental skills were not. The fact that parental support has been associated with protective factors against OSA has already been mentioned previously ( Aljuboori et al., 2021; Jonsson et al., 2019), which highlights the complementary importance of both strategies of support such as parental supervision ( Livingstone Helsper, 2008).
The third objective of this study was to assess how parental profiles of support, supervision, and Internet skills are associated with protective and risk factors. The results recognize three profiles of parents, all of them with similar competence on the Internet, but who differed in the support and supervision of their children: balanced (support and supervision), controllers (supervision but no Figure support), and negligent (no supervision, nor support). These three groups are consistent with the traditional classification that divides parents into democratic, authoritarian, and permissive ( Baumrind, 1991). Democratic parents have clear rules, supervise, and monitor while promoting communication and reasoning with their children, which is similar to the characteristics found in this study in balanced parents. Authoritarian parents are similar to what we call controlling parents in this study since they are more oriented towards controlling and supervising, with little disposition for dialogue. Permissive parents are similar to negligent parents, where there would be no greater regulation of their children ( Baumrind, 1991). This is interesting since recent systematic reviews conclude that children with socalled democratic parents are the ones who present better outcomes (Jorge González, 2018; Ruiz-Hernández et al., 2019). The results of this study are consistent with this since the best outcomes were observed in the children of parents who balance control and support when carrying out their role, which supports hypothesis four.
No difference was found between the three parenting styles and the adolescents Internet use frequency. The fact that the adolescents within three groups used the Internet frequently is not surprising, considering the present study was carried out during the COVID-19 pandemic in which most of the social and academic activities of young people were carried out online ( Ponce et al., 2021). However, significant differences were observed in the protective factors. The group with the best results in knowledge of OSA, well-being, and self-efficacy was the adolescents with balanced parents, who combined high support and high supervision. This is highly consistent with previous antecedents showing the importance of balancing control with emotional support ( Milovidov, 2017).
The adolescents with controlling parents (supervi‐sion without support) were the ones who showed less well-being and less self-efficacy, which shows that control in the absence of a supportive emotional climate weakens the personal resources of adolescents and leaves them more vulnerable (Jorge González, 2018) and more in need of seeking affection on the Internet, with the risks that this entails (European Online Grooming Project, 2012; Staksrud, 2013).
Adolescents with negligent parents (without supervision or support) were the ones with the lowest level of knowledge about OSA, which accounts for the scarce involvement of parents in the training process of their children regarding the dangers they face on the Internet. These results confirm what was reported by previous research, that a bad family climate is associated with more risky behaviors (contact with strangers on the Internet) and more online and offline sexual solicitations (Vanwese‐nebeeck et al., 2018; Whittle et al., 2013).
4.1 Limitations and future lines of research
The main limitation of this study is the small sample size. This prevents generalizing the results and carrying out deeper analyses that incorporate, for example, gender differences. Future studies should make efforts to incorporate larger and more representative samples that allow for more complex analyses, incorporating sociodemographic differences (e.g., gender, nationality, ethnicity, rural residence, socioeconomic level).
CONCLUSION
Although the limitations of this study, the results highlight the importance of parents in preventing OSA in adolescents. Although the parental profile is not associated with the frequency of Internet use in adolescents, it is clearly associated with important OSA prevention factors such as well-being, self-efficacy, and adolescents knowledge of the risks of the Internet. This confirms what was stated by UNICEF (2016, 2020) regarding the importance of parents as mediators between adolescents and technology. Parents should be included in prevention strategies because their role can make a difference in essential factors associated with it. These strategies should enable parents to balance the role of control with the role of support. Toro and Guerra (2021) propose to focus on the principle of progressive autonomy, focusing more on control in young children and gradually focusing more on support in older children. However, it may be necessary for parents to move in a dynamic balance between both roles according to their childrens circumstances at a specific time (more controlled when there is more danger and more understanding when the adolescent shows a capacity for self-regulation and self-care). These findings are essential inputs to inform public policies regarding the prevention of OSA. However, it is necessary to continue researching the topic. In particular, it could be relevant to generate training programs for parents to strengthen their protective role and measure the impact this has on their childrens use of the Internet and the risk of suffering from OSA.
CONSENT TO PARTICIPATE
Participants and parents/guardians confirmed their consent to participate by signing a consent form.
CONSENT FOR PUBLICATION
Participants confirmed that they had read and understood the information sheet, which stated that the studys results may be summarized in published articles.
CONFLICT OF INTEREST
The authors declare that they have no conflict of interest
FUNDING
There is no funding source.
INFORMED CONSENT
Informed consent was obtained from all individual participants included in the study.
DATA AVAILABILITY
Data is not available due to limits of informed consent.
ETHICS APPROVAL
This study received ethical approval from the Comité de Ética Centro Norte de la Universidad Santo Tomás, Chile, on 20 December 2020, reference Nº169/2020.
AUTHOR CONTRIBUTIONS
Material preparation and data collection were conducted by Cristóbal Guerra, Elizabeth Swiergala, Claudia Melo, Carolina Alcorce, Cherie González, Angelica Fernández, Geraldine Martínez, Maira Ramírez. Cristobal Guerra, Cristian Pinto-Cortez, and Edgardo Toro performed statistical analysis. All authors read and approved the final manuscript.
ACKNOWLEDGMENT
To Magaly Retamal for her valuable collaboration in the data collection process.