Adolescents’ level of knowledge of and supportive attitudes to sexual crime in the Swedish context

ABSTRACT

The aim of the present study was twofold. First, we wanted to quantify the level of knowledge of Swedish young people regarding sexual crime and to evaluate their supportive attitudes, while at the same time we aimed at identifying, through self-report, the sources that most contribute to such knowledge and attitudes. A sample of 245 upper secondary school students was selected from five schools in four Swedish counties. The results indicate that adolescents in Sweden have a high level of knowledge of rape, sexual molestation/harassment, and sexual exploitation of a dependent person. Furthermore, they show non-supportive attitudes to rape, sexual harassment, and sexual crime in general. However, some issues related to these types of crime proved to be confusing to the participants and, therefore, require targeting in education policies, specifically among juvenile males and those born abroad. The results are discussed in the context of the needs for sexual crime prevention.

KEYWORDS:

• Sexual crime
• crime prevention
• adolescence
• attitudes
• knowledge

During 2015, 18,100 sexual crimes were reported in Sweden (Swedish National Council for Crime Prevention – Brottsförebyggande rådet [Brå], 2017a) where sexual molestation/harassment and rape are the most common types of sexual crime reported each year, followed by sexual coercion and sexual exploitation (Brå, 2015a). The high prevalence most likely, partially, represents the real number of sexual crimes perpetrated, since these are greatly underreported (e.g. Brå, 2015b; Daigle, 2013). A study among Swedish adolescents in year 9 (most of them 15 years old) estimated that only 9% of those who had been sexually coerced and 4.1% of those who had been sexually molested/harassed during the previous 12 months had reported the incident to the police (Brå, 2016). The majority of the victims in Sweden are aged between 16 and 24 years and are females more often than males (Brå, 2017b), which is consistent with age and gender being well-known risk factors in sexual crime victimisation (e.g., Daigle, 2013; Witkowska, 2005; Yee, Alagappar, & Ngeow, 2015). Preventing sexual crime should be a policing priority, since the repercussions for victims are serious and multiple. In the literature, dissatisfaction with jobs or problems in school, problems with friends and family, depression and psychological distress, somatic symptoms, risky sexual behaviour, and risk of future sexual victimisation (e.g. Fineran & Gruber, 2009; French & Neville, 2012; Mitchell, Ybarra, & Korchmaros, 2014; Zetterström Dahlqvist, Landstedt, Young & Gillander Gådin, 2016) have been described among other consequences. One way to prevent sexual crime is to increase the knowledge about it and change supportive attitudes to it. The deterrent effect of criminalisation could depend on this (e.g. Jung, Ahn-Redding, & Allison, 2014; Sarnecki, 2009). Accurate knowledge of sexual crime has proved to be a protective factor against the acceptance of rape myths and the perpetuation of sexual offending (Aronowitz, Lambert, & Davidoff, 2012; Maxwell, Robinson, & Post, 2003) and increasing the report rates (Daigle, 2013). More sexual knowledge, including knowledge of the legalities, leads to less rape-supporting beliefs (Aronowitz et al., 2012; Mallet & Herbé, 2011). Adolescent males who have a good knowledge of rape and reject rape stereotypes are less likely to report engaging in incidents of sexual offending (Maxwell et al., 2003). On the other hand, supportive attitudes to sexual crimes, such as rape myth acceptance, gender role stereotypes, and aggressive sexual attitudes have been linked to sexual offending among juveniles (e.g., Kjellgren, Priebe, Svedin, & Långström, 2010; Lacasse & Mendelson, 2007; Lanier, 2001; Martins, Machado, Abrunhosa, & Manita, 2012; Moyano, Monge, & Sierra, 2017; Shen, Chiu, & Gao, 2012; Zakireh, Ronis, & Knight, 2008). Knowledge of the law is important to be able to make sound decisions regarding what an individual decides to support personally (Jung et al., 2014). It is important to analyse attitudes in order to develop interventions and policies (Levin & Peled, 2011). Adolescence might be a golden period for enhancing knowledge and promoting less sexual crime supportive attitudes. It is a period of transition in which adolescents feel pressured to have their first intimate relationships (Lacasse & Mendelson, 2007). Ignorance, confusion, and lack of experience can originate behaviours of sexual coercion among non-skilled adolescents and make victims unaware of their condition. Behaviours learnt and attitudes developed during adolescence are likely to be carried on in adult life (Sanchez, Fetterolf, & Rudman, 2012; Sherer, 2010; Smith, White, & Holland, 2003).

The scientific literature provides mixed results about juveniles’ knowledge of sexual crime. Lack of knowledge has been reported in many ways in different studies in the past. When presented by researchers with various scenarios, adolescents showed difficulties in recognising sexual crime other than child molestation and rape after petting (Lewin, 1983). Incest and child sexual abuse was scarcely identified as non-consensual sexual behaviour (Jordan, Price, Telljohann, & Chesney, 1998). Verbal harassment and forced unprotected sexual intercourse was not identified as violent criminal behaviour (Ohnishi et al., 2011). Young people showed confusion about those behaviours that constitute sexual coercion (Birungi, Nabembezi, Kiwanuka, Ybarra, & Bull, 2011) or who to blame in rape scenarios (Telljohann, Price, Summers, Everett, & Casler, 1995). Stevenson, Najdowski, and Wiley (2013) found that almost half of their juvenile participants did not know that a person under 18 years old can be registered as a sex offender, and they were unaware of which behaviours could lead to registration. However, more positive results about adolescents’ knowledge of this type of crime have recently been reported. In a study, in which behaviours constituting rape were evaluated through a series of seven rape myth statements, the majority of high school students answered five or more questions correctly (Lee, Stark, O’Riordan, & Lazebnik, 2015). In other studies, high school students were able to respond correctly to questions about sexual assault (Daigneault et al., 2015) and sexual harassment (Witkowska, 2005). Ethnicity and sex are variables that proved to have an impact on both negative and positive results. In general, those young people of European heritage have a better knowledge than those from other ethnic backgrounds (Jordan et al., 1998) as well as females when compared to males (Jordan et al., 1998; Lee et al., 2015; Ohnishi et al., 2011; Witkowska, 2005). Higher knowledge among females might be explained by a higher probability of having experienced, witnessed, and been raised to be aware of sexual violence (Breiding et al., 2014; Drennan, Hyde, & Howlett, 2009; Lee et al., 2015; Otwombe et al., 2015).

In relation to adolescents’ attitudes, tolerance of rape myths (e.g. Aronowitz et al., 2012; Kershner, 1996; Tavrow, Withers, Obbuyi, Omollo, & Wu, 2013; Waubert de Puiseau & Roessel, 2013), rape (Telljohann et al., 1995), sexual assault (Daigneault et al., 2015; Davis & Lee, 1996), dating violence (Geiger, Fischer, & Esheet, 2004; Shen et al., 2012; Sherer, 2010) and conservative attitudes to gender roles (e.g. Epps, Haworth, & Swaffer, 1993; Foulis & McCabe, 1997; Kershner, 1996) have been identified in the literature. Young people were characterised as having a high acceptance regarding sexual harassment (Foulis & McCabe, 1997) and as holding stereotypes regarding the justification of sexual coercion in dating situations (Aronowitz et al., 2012; Geiger et al., 2004). Jordan et al. (1998) found more positive results in that a low percentage of juveniles in their study agreed with supportive statements on sexual harassment, and Witkowska (2005) found that Swedish adolescents considered it a problem. Sex, age and ethnicity seem to have an impact on adolescents’ attitudes. Males hold more supportive beliefs on and attitudes to sexual crime than females (Davis & Lee, 1996; Geiger et al., 2004; Jordan et al., 1998; Mallet & Herbé, 2011; Maxwell et al., 2003; Strouse, Goodwin, & Roscoe, 1994; Witkowska, 2005). Younger individuals have been stated to endorse more “adversarial sexual attitudes toward women” (p. 264) than older ones (Emmers-Sommer, 2015), and Caucasians (Jordan et al., 1998) and those of a Jewish background (Sherer, 2010) show less supportive attitudes than other ethnicities. Ethnicity is a controversial factor that is evaluated in different ways by different authors, but always shows a relationship with attitudes and tolerance to sexual crime. Most likely ethnic groups differ in their sexual values considering the diverse cultural, political and religious influences on gender role norm endorsement (Safdar & Kosakowska-Berezecka, 2015), and consequently on sexuality (Ahrold & Meston, 2010; Amaro, Navarro, Conron, Ray, & On, 2002; Leiblum, Wiegel, & Brickle, 2003). In more gender egalitarian countries, women may experience subtle forms of discrimination, but in societies with greater gender inequality discrimination and hostility towards women are still openly expressed (Vescio, Gervais, Snyder, & Hoover, 2005). Therefore, the attitudes to and the knowledge of what is constitutive of a sexual crime must be targets to explore among different ethnic groups, mainly in diverse societies, as it is Sweden where those born abroad represent almost 15% of the population in the country (Statistics Sweden – Statistiska centralbyrån, 2016).

Apart from age, sex and ethnicity, several other factors are associated with the level of knowledge of and attitudes to sexual crime among adolescents. Several studies demonstrated the impact of the family (Fehler-Cabral & Campbell, 2013; Martins et al., 2012; Strouse et al., 1994), friends (Fehler-Cabral & Campbell, 2013; Martins et al., 2012; Yee et al., 2015), teachers (Fehler-Cabral & Campbell, 2013; Owens, Shute, & Slee, 2005), media along with advertisements (Brown & L’Engle, 2009; Emmers-Sommer, Pauley, Hanzal, & Triplett, 2006; Witkowska, 2005), internet (Brown & L’Engle, 2009) and film, tv or music (Brown & L’Engle, 2009; Emmers-Sommer et al., 2006; Strouse et al., 1994), which we decided to evaluate in our research too. With this exploratory study, we aimed at identifying the need to introduce programmes in schools to increase awareness and sensitise adolescents to sexual crime victimisation and perpetration in the Swedish context. Many studies have found different intervention programmes in schools to be effective in improving adolescents’ knowledge and attitudes (e.g. Daigneault et al., 2015; de Lijster, Felten, Kok, & Kocken, 2016; Lundgren & Amin, 2015), although students and teachers have been stated to have different perceptions regarding how helpful interventions might be (e.g. Jordan et al., 1998; Owens et al., 2005). Two specific objectives led our study: (1) to quantify adolescents’ level of knowledge to and evaluate supportive attitudes to sexual crime on the basis of gender, age and immigrant background and (2) to identify the sources that most contribute to adolescents’ knowledge and supportive attitudes to this specific type of crime.

Method

The study had a cross-sectional design, based on a purposively developed paper questionnaire.

Participants

A non-probabilistic sample of 245 high school students aged 16–19 years (M = 17.2, SD = 1.0), who accepted voluntarily to participate in the study, was selected from five schools in four Swedish counties (Södermanland, Gävleborg, Västernorrland, and Norrbotten). Of the 245 participants, 241 reported their gender as being 104 (43.2%) males, 134 (55.6%) females, and 3 (1.2%) identified themselves as being of “Other” gender. Two hundred and twenty-six participants disclosed their age, of whom 73 (32.3%) were 16 years old, 67 (29.6%) were 17, and 86 (38.1%) were 18 or 19. Country of birth was disclosed by 227, where 176 (77.5%) were born in Sweden and 51 (22.5%) were born abroad.

Measures

The questionnaire was distributed to be completed in the classroom and included four parts: (1) demographic variables (i.e. gender, age and immigrant background), (2) measurement of the level of knowledge through scenarios, (3) measurement of supportive attitudes through scale-rated items, (4) sources of knowledge and attitudes.

Knowledge of sexual crime

The level of knowledge was assessed using 12 short scenarios depicting different sexual situations (see Table 1 in Supplemental on-line material). Participants had to state whether they believed the situation was a sexual crime or not using the alternatives “Yes”, “No” or “I don’t know”. The use of scenarios was chosen because the employment of fictitious people creates a distance between the subject and the respondent, which elicits more accurate responses since the participant will not feel threatened by the question or be as inclined to give a socially desirable answer (Finch, 1987). However, it has been suggested that the reference to specific individuals (vs. to abstract groups) increases the desire to punish on the part of research participants (Salerno et al., 2010), which might play a role in making participants more prone to consider the specific situations displayed in the scenarios as crimes. The scenarios were composed to illustrate sexual crimes according to Swedish law (SFS, 1962:700). Since the law comprises many different crimes, a few were chosen – (1) rape, (2) sexual molestation/harassment, and (3) sexual exploitation of a dependent person – based on their relevance to adolescents and their higher prevalence in Sweden compared to other sexual crimes. Sexual content non-criminal act scenarios were also included so as to avert biased one-tailed answers (e.g. always “Yes”). The language of the scenarios was adapted in order to be appropriate to the age group and represented persons close in age in order to enhance the relevance to them. Nine of the scenarios illustrated a sexual crime – e.g. “Daniel, 18 years old, gets so drunk at a party that he falls asleep on a couch. While he is sleeping Andreas, 19 years old, pulls down his pants and gives him a blow job”, while three of the scenarios depict a non-criminal sexual situation – e.g. “Erik and Maja, both 15 years old, have sex one night during the summer holiday. Both agreed to the intercourse”.

Attitudes to sexual crime

Thirty-one statements were used so as to measure supportive attitudes to rape, sexual harassment, and sexual crime in general (see Table 2 in Supplemental on-line material). To measure supportive attitudes to rape, seven statements from the Rape Myth Acceptance Scale (RMA; Burt, 1980) were adapted in order to make them age consistent – e.g., “If a girl wears a short skirt and a tight top she only has herself to blame if something happens”. Following the same age consistency reasoning, 15 items from the Sexual Harassment Attitude Scale (SHAS; Mazer & Percival, 1989) were adapted in order to measure supportive attitudes to sexual harassment – e.g. “One of the problems with sexual harassment is that boys can’t take a joke”. Finally, general supportive attitudes to sexual crime were measured through nine items – e.g. “Anyone who is exposed to a sexual crime while they are drunk is partially responsible for what happened”. Eight of these items were adapted from a survey about community attitudes regarding sexual assault by Massachusetts Institute of Technology (MIT, 2014). One item was inspired by other sources (Foshee, Fothergill, & Stuart, 1992; World Health Organization, 2017). Statements were gathered in three clusters since attitudes to sexual crime may differ depending on the type of crime (Kennedy & Gorzalka, 2002).

The participants were asked to rate their agreement with each statement on a 7-point Likert scale, where 1 represented “Strongly Disagree” and 7 “Strongly Agree”. The items of each one of the crime clusters were summed to achieve a cluster-attitude composite score. The clusters showed moderate-good internal consistency values, with α = .78, .81, and .84 for rape, sexual harassment and general attitudes to sexual crime respectively.

Sources of influence on the level of knowledge and attitudes to sexual crime

The sources that possibly have an impact on the level of knowledge and supportive attitudes to sexual crime were asked for through two independent questions which had multi-optional but not mutually exclusive answers, since the alternatives could overlap. Six fixed alternatives were chosen: “Family”, “Friends”, “Teachers”, “Media or advertisements”, “Internet”, and “Film, TV or music”. A final alternative, “other”, was added in order to give the participants an opportunity to answer the question when none of the above-mentioned alternatives was appropriate. In such case, participants were requested to make the source explicit. Books, courses, culture, experiences, myself, common sense, newspapers, nurses, the police and authorities, the law, and documentaries were sources stated by different participants.

Procedure

During March 2017 school administrations in several counties in Sweden were contacted by phone. After being given a brief presentation of the study’s objectives and methods, five schools agreed to collaborate. Teachers were then contacted and agreed on a day and time to distribute the questionnaire in classroom. Before the sessions of data collection, students were informed orally and in writing about the study, of the fact that their answers would be anonymous, and that their participation was voluntary. Information sheets were distributed among all students who were given the chance to ask questions about their participation. It was clear that consent to participate was considered to be given by returning the questionnaire. Since the questionnaire was anonymous, students could not withdraw consent afterwards. Researchers’ contact information was included in the information sheet together with the telephone number of the school’s student health personnel, a victim support group working in the area and the police in case they felt they needed to talk to someone after reading or completing the questionnaire. All the students who were present in the classrooms on the days the questionnaire was distributed chose to participate. Teachers were mostly present but were not actively involved in the completion of the questionnaire

Statistical analysis

The data was analysed using IBM SPSS Statistics Version 24 for Windows. Three participants identify themselves with a gender different from male or female, so they were left out of the gender analyses. Age was analysed in three groups; 16, 17 and 18–19. Analyses related to an immigrant background were performed on the basis of country of birth (i.e. Sweden vs. abroad).

When analysing scenarios, one point was given to each scenario answered correctly, giving a range of possible composite score from 0 to 12. A high score indicated a good knowledge of sexual crime. If a participant had only answered 11 scenarios, the composite score was prorated. The score of participants who answered 10 or fewer scenarios was considered not valid and was therefore excluded, although participants were taken into account for the attitudes’ analyses in case they had provided valid answers on that part of the questionnaire. The mean composite score of the scenarios (i.e. level of knowledge) was compared between groups of gender, country of birth, and age. To assess the magnitude and practical importance of comparative results in which p < .05, effect sizes were calculated using eta squared.

Supportive attitudes to sexual crime were analysed on the basis of the three item-clusters; (1) rape, (2) sexual harassment, and (3) general attitudes to sexual crime. The three clusters yielded three different composite score rankings. For rape, the rank was between 7 and 49; for sexual harassment, it was between 15 and 105; and for general attitudes between 9 and 63. The range for the total composite score (i.e. the sum of the three cluster composite scores) was between 31 and 217. A higher score indicated more supportive and, therefore, more problematic attitudes to sexual crime. Scores were prorated when participants answered 6 (from a total of 7) items on rape, 13 or 14 (from a total of 15) items on sexual harassment, and 8 (from a total of 9) items of general attitudes. If participants had answered fewer than these thresholds, the answers were considered not valid and therefore excluded from the analyses. Being excluded from one composite score did not necessitate being excluded from the other two, since these could still be valid. Conversely, being excluded from one composite score meant being excluded from the total composite score.

Demographic groups were compared using t-test and ANOVA. A Kruskal–Wallis test was employed to compare the attitudes of the three age groups, since the assumption of homogeneity of variances was violated. One-step linear multiple regression was calculated so as to measure the contribution of each one of the studied sources for the adolescents’ level of knowledge and supportive attitudes. For attitudes, the total composite score (i.e. the sum of the three individual cluster composite scores) was used as an outcome variable in the regression model.

Results

Knowledge of sexual crime

With a possible composite score of 12 on the scenarios part of the questionnaire, the mean value for the whole sample was 8.8 (SD = 1.8), indicating that the participants had a level of knowledge situated on the top third score’s scale. Most often correctly answered were scenario #3 illustrating two 15-year olds having consensual intercourse (a non-criminal act) and scenario #9 depicting a boyfriend forcing his girlfriend to have sex with him by physical means (rape), which were correctly answered by 92.2% and 95.1% of the adolescents, respectively (See Table 1 in Supplemental on-line material). At least 86.4% of the sample answered scenario #4 correctly, illustrating a young man who falls asleep drunk and receives oral sex without his consent (rape), scenario #6, depicting a girl having to masturbate her boss in order to get a day off from work (sexual exploitation of a dependent person), and scenario #10, characterising a boy who fondles a girl on her behind and breasts and calls her a whore (sexual molestation/harassment). On the other hand, some of the scenarios proved to be less knowledgeable by the participants. These were scenario #5 depicting a girl receiving a picture of a penis (sexual molestation/harassment), scenario #7 depicting a man flashing his genitals at a young girl (sexual molestation/harassment), scenario #8 depicting an employee offering sexual services to her boss in exchange for a pay rise (a non-criminal act in the field of sexual exploitation of a dependent person) and scenario #12 where a boy fondles a girls’ behind in school (a non-criminal act in the field of sexual molestation/harassment). None of these scenarios were answered correctly by more than 65.2% of the sample and at least 24.2% of the participants who answered stated that they did not know whether the scenario represented a sexual crime or not. Differences were found among groups on the basis of their gender, country of birth and age (See Table 1).

Table 1. Group comparison – the level of knowledge of sexual crime (scenarios’ composite score).

	n^a	Mean (SD)	CT^b (p value)	η^2
Gender	237
Boys	103	8.5 (1.8)	−2.373 (.018)	.02
Girls	134	9.1 (1.7)
Country of birth	227
Sweden	176	9.1 (1.6)	2.841 (.005)	.03
Abroad	51	8.3 (1.8)
Age	225
16	73	9.3 (1.4)	4.432 (.013)	.04
17	67	8.5 (1.8)
18–19	85	8.8 (1.8)

^aNumber of participants with a valid composite score for knowledge of sexual crime.

^bCT: comparison test. Comparison of gender and country of birth was obtained through t-test. Age groups were compared using ANOVA.

Females show a higher level of knowledge than males, as well as those adolescents born in Sweden in comparison with those born abroad. A difference in age was found as well, with the post-hoc test indicating a better knowledge on the part of the 16-year-old group compared to the 17-year olds (Mean difference (I-J) = 0.81; p = .012). However, the small effect sizes according to Cohen’s criteria (1988) obtained in these comparisons indicates that the differences were of low magnitude and therefore of little practical importance.

Attitudes to sexual crime

The mean of composite scores for all the three clusters of statements were located in the lower range of possible scores, indicating that the adolescents in our sample have low supportive attitudes to rape (M = 15.0, SD = 8.3), sexual harassment (M = 38.0, SD = 13.5), and sexual crime in general (M = 20.4, SD = 10.4). The mean of the total composite score was, of course, low (M = 73.1, SD = 28.8).

Among the supportive attitudes to individual statements of rape, the one suggesting that a girl who follows a young man home after their first date shows that she is willing to have sex received the highest mean score (i.e. highest support), while the one suggesting that a girl who wears revealing clothes only has herself to blame if something happens received the lowest mean score (i.e. lower support) (see Table 2 in Supplemental on-line material). Regarding the individual statements about sexual harassment, the statement proposing that sexism and sexual harassment are two totally different things received the highest support, while the statement suggesting that it is okay for a teacher to accept an offer of sexual services from a student for grade-enhancing purposes obtained the lowest support. Among the individual statements about general attitudes to sexual crime, the statement suggesting that a sexual incident can only be seen as an assault if the person says “no” received the highest support, while the statement proposing that it is okay for someone to force their partner to have sex obtained the lowest support. The mean composite scores for each group, in each one of the three clusters of statements on attitudes to sexual crime, are presented in Table 2.

Table 2. Group comparison – attitudes to sexual crime (items’ composite scores).

	Rape				Sexual harassment				General attitudes
	n^a	Mean (SD)	CT^b (p value)	η^2	n^c	Mean (SD)	CT^b (p value)	η^2	n^d	Mean (SD)	CT^b (p value)	η^2
Gender	228				204				217
Boys	101	18.9 (9.3)			89	44.2 (13.3)			91	25.3 (10.6)
Girls	127	11.7 (5.8)	6.827 (.000)	.17	115	32.7 (11.0)	6.592 (.000)	.18	126	16.7 (8.4)	6.369 (.000)	.16
Country of birth	219				201				213
Sweden	174	12.5 (6.1)			163	35.0 (11.7)			173	18.2 (8.8)
Abroad	45	20.9 (9.4)	−5.710 (.000)	.13	38	47.5 (13.6)	−5.733 (.000)	.14	40	28.4 (11.9)	−5.124 (.000)	.11
Age	218				198				211
16	72	12.7 (6.2)			66	36.3 (9.7)			70	18.3 (7.4)
17	63	14.4 (7.7)			57	37.3 (13.4)			63	18.8 (9.8)
18–19	83	16.8 (9.6)	5.639 (.060)	–	75	38.6 (15.5)	0.047 (.977)	–	78	22.5 (12.4)	3.510 (.173)	–

^aNumber of participants with a valid composite score for attitudes to rape.

^bCT: comparison test. Gender and country of birth groups were compared using t-test, age groups was compared using Kruskal–Wallis test.

^cNumber of participants with a valid composite score for attitudes to sexual harassment.

^dNumber of participants with a valid composite score for general attitudes to sexual crime.

Statistically significant differences were found among gender and country of birth groups with effect sizes indicating a moderate result, and therefore one of practical significance. The biggest effect sizes were found for the composite scores related to sexual harassment (η² = .18, .14 for gender and country of birth, respectively) closed followed by rape (η² = .17, .13). Although lower effect sizes were found for general attitudes to sexual crime (η² = .16, .11), they are still at a moderate level following Cohen’s criteria. Specifically, males and those adolescents who were born abroad have more supportive, and therefore problematic, attitudes to the three clusters of sexual crime. In relation to age, no differences were found among the three groups.

Sources affecting the level of knowledge of and attitudes to sexual crime

As can be seen in Table 3, the alternative which was most often stated by the adolescents to impact their level of knowledge was the “Internet”, followed by “Media or advertisements”. “Friends” was also pointed out by half of the participants, while those sources which are supposed to have an educative role (i.e. “Family” and “Teachers”) were less often stated. However, when all the sources were introduced on a linear multiple regression, “Family” showed to have the steepest positive slope (β = .184, p < .000), followed by “Media or advertisements” and “Internet” (β = .174, p = .011, and β = .159, p = .017, respectively), while unexpectedly “Friends” and “Teachers” proved to be not significantly associated with the level of knowledge.

Table 3. One-step multivariate linear regression to measure the impact of different sources on the level knowledge of and supportive attitudes to sexual crime.

	Knowledge (n = 238)			Attitudes (n = 204)
Predictors	Yes – n (%)	β (p value)	Adjusted R^2	Yes – n (%)	β (p value)	Adjusted R^2
Family	107 (45.0)	.184 (.006)		116 (56.9)	−.273 (.000)
Friends	123 (51.7)	.019 (.780)		162 (79.4)	−.089 (.167)
Teachers	111 (46.6)	.032 (.598)		76 (37.3)	−.089 (.187)
Media, Advertisements	154 (64.7)	.174 (.011)		142 (69.6)	−.168 (.011)
Internet	186 (78.2)	.159 (.017)		152 (74.5)	−.211 (.002)
Film, TV, Music	88 (37.0)	−.054 (.420)		81 (39.7)	.070 (.281)
Other	19 (8.0)	−.016 (.800)	.089	19 (9.3)	−.216 (.001)	.237

In relation to attitudes to sexual crime, “Friends” was the source most frequently indicated, followed closely by “Internet” and “Media or advertisements”. Once again, “Family” and “Teachers” were less often pointed out. Similarly, in the result on knowledge, the linear multiple regression, revealed that “Family” was the strongest (negatively) associated source to attitudes (β = −.273, p < .000), meaning that they have the greatest influence on less supportive attitudes. “Other” sources also showed a (negative) steep slope (β = −.216, p = .001) followed by “Internet” and “Media or advertisements” (β = −.211, p = .002, and β = −.168, p = .011, respectively) which might have important significance for prevention.

Discussion

While most past research carried out on both adolescent and adult samples shows a low level of knowledge about sexual crime (e.g. Birungi et al., 2011; Crisanti, Arboleda-Florez, & Stuart, 2000; Jordan et al., 1998; Jung et al., 2014; Lee et al., 2015; Lewin, 1983; Roberts, Grossman, & Gebotys, 1996; Stevenson et al., 2013), a general appraisal of the results of our study indicates that the great majority of adolescents in Swedish schools classified correctly (crime vs. non-criminal act) scenarios of the rape, sexual molestation/harassment, and sexual exploitation of a dependent person. Only non-criminal scenarios showed a less correct percentage of answers. The finding is not new though, since recent studies in juvenile populations, within and outside of Sweden, have found similar results (e.g. Daigneault et al., 2015; Lee et al., 2015; Witkowska, 2005). It is noteworthy, however, that some scenarios were less frequently correctly answered by the participants and these therefore constitute areas that should be targeted for active debate with the young people in order to prevent sexual criminal behaviour and victimisation. Specifically, on the sexual molestation/harassment issue, the reception of unwanted sexually explicit images through the Internet – very important since adolescents are often connected and are therefore exposed for many hours of the day – and sexual assault, consisting on rubbing the genitals against a non-consenting person, seemed to generate confusion as to whether they constitute a crime or not, compared with rape scenarios. One interpretation might be that it is easier to recognise the substantial content of a rape situation than one of a sexual molestation/harassment. Roberts et al. (1996) found that adults often knew about the core parts of the Canadian rape law reform even if their knowledge of sexual crime in general was low. The perceived seriousness of a behaviour affects whether it is regarded as a crime or not (Herzog & Einat, 2012; Houston & Hwang, 1996), and rape is most likely perceived as being more serious than other sexual criminal behaviours. This might also explain why, in our study, rape in a relationship context was more easily identified as a sexual crime than receiving a picture of a penis. On the other hand, language has been proven to impact on research when research participants are asked about sexual violence (McKenzie-Mohr & Lafrance, 2011). In our study, the language used to describe each scenario might hold some clues about whether the act was malicious (e.g. scenario #9 states that a boyfriend force his girlfriend to have sex and the word “force” has negative connotation) or might have led to a certain level of confusion (e.g. scenario #5 states that a girl receives a picture which is in a passive form) and therefore influenced participants’ answers.

Adolescent females proved to have better knowledge than adolescent males (Ohnishi et al., 2011; Yee et al., 2015). In our study, this difference is, however, of little practical significance. Therefore, we consider that interventions designed to improve adolescents’ knowledge of sexual crime should target both sexes in the same way. Differences of little practical importance were also found in relation to the immigrant background. Since Sweden has, for many years, been a country receiving refugees and a workforce immigrant population from many countries with different ethnical backgrounds – most likely entailing some values and rules different from those common in the Swedish society – we expected to find greater differences at the level of knowledge on the basis of country of birth. Results pointed to the fact that those adolescents who were born in Sweden have a slightly higher level of knowledge than those born abroad, but the differences were rather small. So, interventions to raise the level of knowledge must be taken without distinction of immigrant background.

The less accepting attitudes to sexual harassment among the participants in our study are consistent with those presented by Witkowska (2005) in her study of Swedish adolescents – despite the 12-year difference between the studies and the changes in both schools and society during that time. It can, perhaps, in part explain the adolescents’ good knowledge of sexual crime according to the suggestion that less tolerant attitudes lead to the perception of more deviant sexual behaviours as a crime (Shechory Bitton & Ben Shaul, 2013). Furthermore, young Swedish people today are perhaps more conscious of sexual crime due to the attention given to the issue in the media (Brå, 2015a) and general trends in society (e.g. gender equality). However, there are still areas of concern to work out in prevention. The previous reported results by Aronowitz et al. (2012) and Geiger et al. (2004) regarding sexual advances being something to expect in dating situations was somehow replicated in this study with the high support for the statement making the assumption that following someone home after a date means willingness to have sex.

Although demographic groups were not notably different in their level of knowledge, we found significant differences in attitudes that should be taken into account for any intervention programme. Specifically, boys and those adolescents who were born out of Sweden showed more supportive (and therefore problematical) attitudes especially as regarded sexual harassment followed closely by rape. These groups should be prioritised in prevention specifically targeting a change of attitude. The sex differences found with females having less supportive attitudes to sexual crime reflects the majority of the results in the literature (e.g. Davis & Lee, 1996; Geiger et al., 2004; Jordan et al., 1998; Mallet & Herbé, 2011; Maxwell et al., 2003; Sherer, 2010; Strouse et al., 1994; Witkowska, 2005). Females are more likely to have experienced (Breiding et al., 2014; Drennan et al., 2009), witnessed (Otwombe et al., 2015) and been raised to be aware of (Lee et al., 2015) sexual violence from males than the other way around. Furthermore, gender role stereotype and socialisation of males might explain the differences as well, since adolescents have to conform to different kinds of factors. The leading norm in many societies is that males are supposed to be less sensitive to this kind of subject, while females are expected to be the opposite.

Although a different birth country does not necessarily imply a different ethnicity, higher support for attitudes to sexual crime found among the foreign-born group in our study was expected, on the basis of ethnic differences found in other studies (Jordan et al., 1998; Sherer, 2010). Segregation might help to explain these differences. Some groups of juveniles born in other countries may not get the same opportunity or find it hard to become part of Swedish society and therefore socialise more within their own group (Mollenhorst, Edling, & Rydgren, 2017; Oppedal, Røysamb, & Heyerdahl, 2005). Since there are differences in sexual crime legislation between different countries, this might make it hard for those born abroad to keep track of what counts as sexual crime according to Swedish law. When these young people do not get involved in society as much as they would like or should, they might retain their old views (Bhugra & Becker, 2005; Phinney, Horenczyk, Liebkind, & Vedder, 2001). Furthermore, different cultures will probably have different definitions of acceptable behaviour (Akers, 1973, 1998), and attitudes to sexual harassment could be mostly influenced by culture. In relation to age, we did not find significant differences among groups of 16, 17, and 18 or 19 year-olds, indicating that attitudes are likely developed during early adolescence, as proposed by several authors (Emmers-Sommer, 2015; Sanchez et al., 2012; Sherer, 2010; Smith et al., 2003). Moreover, gender differences on attitudes to rape and sexual harassment found in adult populations (e.g. Jimenez & Abreu, 2003; Kennedy & Gorzalka, 2002; Shechory Bitton & Ben Shaul, 2013; Sierra, Santos-Iglesias, Gutiérrez-Quintanilla, Paz Bermúdez, & Buela-Casal, 2010; Talbot, Neill, & Rankin, 2010) are according to the male vs. female differences found in our sample of adolescents adding credence to the hypothesis that these types of attitudes developed at an early age. This reinforces the need to develop intervention programmes at early ages.

Regarding the sources of knowledge and attitudes, family, as expected, proved to play an important role. Family most often is the adolescent’s main network apart from friends, in normalised populations, until they start seeking independence (Thornberry & Krohn, 2005), which may not have happened yet, and family could remain an important part of their network, especially when they still live at home. The family is the adolescent’s primary source of observation and feedback, forming their opinions of sexuality and in this way increasing adolescents’ level of knowledge and enhancing their non-supportive attitudes. The fact that teachers were not indicated as a source of knowledge nor important for the development of attitudes to sexual crime was an unexpected and concerning result. Schools, together with families, should work together to shape knowledge and attitudes. Different aspects of sexuality as well as sexual crime should be clearly taught and openly discussed in schools. Teachers, together with parents, should be a primary source of knowledge not only in maths and language but also in matters of such social relevance as sexuality and sexual crime. Although sexual health, reproduction, and sexual orientation are topics contemplated in the Swedish elementary and high school curricula (Skolverket, 2013a, 2013b, 2014), they do not include topics related to sexual violence or sexual crime. We think that it is of importance that education policies should prioritise this matter of introducing specific programmes in schools. As concerns sources other than family or teachers, the media together with advertisements were identified as sources with impact, which was already found by Witkowska (2005) 12 years before. Media or advertisements as well as the Internet could be channels by which the adolescents connect with others to create new networks (Thornberry & Krohn, 2005). Through these channels, they observe and receive feedback about different matters and behaviours. These sources should be very closely monitored because, although they seem to be working in a proper way (i.e. producing more knowledge and less supportive attitudes to sexual crime), there is no control over the traffic on the net and the opposite effect (i.e. misleading knowledge and highly supportive attitudes to sexual crime) might easily occur. That “Other” sources have an impact on the type of attitudes the adolescents show is of concern as well, since there is little control of them. Some of these factors are centred on the adolescents’ own characteristics while others are centred on the factors in the general society. This may indicate that the adolescents stating conscience, experiences and myself might think that they have not been influenced by others in the manner described by social learning (Akers, 1973) or interaction (Thornberry & Krohn, 2005) theories. Instead, they believe themselves to be responsible for creating their own attitudes. Those who stated news, the police and society as affecting their attitudes to sexual crime might be more in line with the above-mentioned theories, since these are sources they may interact with and observe in their daily life. Through these, young people’s attitudes can be changed, and that can occur in both directions (i.e. low or high support for sexual crime). Enhancing the knowledge and promoting less supportive attitudes in young people should then be ensured through all possible sources.

Limitations, strengths and further research

The results of this study should be seen in the light of several limitations and strengths. First of all, we could not gain access to a random sample of adolescents. Therefore, it is not possible to draw a conclusion about the whole population of Swedish adolescents and the results are preliminary in nature. However, participants were from four different counties, which means the results were not tied to a single location, having some geographical spread instead.

Secondly, the scenarios described three types of sexual crime only. Further research should address other acts considered to be criminal in Swedish law. A deeper understanding of the impact of ethnicity, and related variables such as religion, is also a matter for future studies, which could consider variables other than country of birth.

A better knowledge of the sources and how they impact the level of knowledge and attitudes is also required. For example, in the area of sexual crime, the type of family, the kind and quality of family relationships, and how parents provide knowledge and shape attitudes are issues of importance to conduct research on. One important factor to explore in new research is how personal experiences of sexual crime affects the adolescents’ knowledge of and attitudes to sexual crime. In our study, this could be the case where children cited themselves as their own source of knowledge. Gathering information on crime victimisation and perpetration jointly with attitudes and knowledge could produce results of great interest.

As a strength, the study treats a current and important topic in Swedish society, which was agreed by the participants and most likely made them more committed to answer the questionnaire, perhaps more truthfully than if they had not been interested in the subject. However, the study relies on self-reporting, which always includes the risk of socially desirable answers (Bryman, 2008). Knowledge and attitudes to sexual crime has been little researched among juveniles in Nordic countries, and we consider that this study helped to fill a gap and with some new insights into the subject.

Disclosure Statement

No potential conflict of interest was reported by the authors.

ORCID

Teresa Silva http://orcid.org/0000-0002-3955-7688