Study design and setting

The SHoT2018 study (Students’ Health and Wellbeing Study) is a national student survey for higher education in Norway initiated by the three largest welfare organisations (Sammen (Bergen and surrounding area), SiT (Trondheim and surrounding area) and SiO (Oslo and Akershus)). The SHoT2018 was collected electronically through a web-based platform. Details of the study have been published elsewhere,19 but in short the SHoT2018 was conducted between 6 February and 5 April 2018, and invited all full-time Norwegian students pursuing higher education (both in Norway and abroad).

Participants

All full-time Norwegian students aged between 18 and 35 taking higher education (both in Norway and abroad) received both an email and SMS (short message service) text invitation to take part in the study. For the SHoT2018 study, 162 512 students fulfilled the inclusion criteria, of whom 50 054 students completed the online questionnaires, yielding a response rate of 30.8%.

Instruments

Statistics

IBM SPSS V.25 for Mac was used for all analyses.

Negative binomial regression analyses were used to examine gender differences in the prevalence of sexual harassment forms across age groups. Negative binomial regressions (producing incidence rate ratios (IRR)) were used rather than the more commonly used logistic regressions (producing OR), as ORs tend to overstate an effect size compared with IRRs for high frequency outcomes (in this case relevant for some forms of sexual harassment).20 Missing values were handled using listwise deletion.

Patient and public involvement

The planning and design of the study were initiated and governed by the student welfare organisations, which included deciding the inclusion and exclusion criteria, and selecting potential research questions and instruments. Students were not involved in the actual collection of data, although recruitment was conducted in close collaboration with all the student welfare organisations in Norway. The results will be disseminated to the study participants via outlets of the student welfare organisations and educational institutions, with newsletters highlighting the main findings being made available to all students. Popular summaries of results and interpretations with interest for a wider audience will be disseminated in appropriate outlets (eg, the web pages of educational institutions and the Norwegian Institute of Public Health (NIPH)).

Descriptive characteristics and representativeness

Compared with all invited students (58.1% women (n=93 267) and 41.9% men (n=67 558)), the current sample included a larger proportion of women (69.1%) than men (30.9%). The mean age was 23.2 (SD=3.3), and the age distribution of participating students (18–20 years (18%, n=8832), 21–22 years (31%, n=15 471), 23–25 years (32%, n=15 902), 26–28 years (12%, n=5710) and 29–35 years (7%, n=3427)) was almost identical to that of all invited students (18–20 years (18%, n=28 996), 21–22 years (31%, n=49 731), 23–25 years (32%, n=51 714), 26–28 years (12%, n=19 901) and 29–35 years (6%, n=10 216)). In terms of accommodation status, 19.3% reported living alone, while 49.9% reported being single. Eight per cent of the sample were immigrants, defined as either the student or his/her parents being born outside Norway.

Prevalence of sexual harassment

Lifetime sexual harassment was reported by 24.2% of the students, and 16.7% reported having been sexually harassed within the past year. Women reported substantially more sexual harassment than men (lifetime: 31.3% vs 8.0%, IRR=3.9, 95% CI 3.7 to 4.1). A similar gender effect was also observed for past-year sexual harassment (21.6% vs 5.7%, IRR=3.8, 95% CI 3.5 to 4.0; see table 1 for details). The most common forms of lifetime sexual harassment were sexual expressions and suggestions, comments about your body, appearance or private life, and unwanted touching, hugging or kissing, which both were reported by 15.4% of the students. Intrusive eye or body movements were also common (11.4%), whereas viewing/sending sexual images (3.7%) and indecent exposure (1.7%) were less frequently reported. Rape and rape attempt were reported by 3.4% and 2.1%, respectively. As detailed in table 1, large gender differences were observed across all variables.

Figure 1 depicts students’ detailed responses when he/she were last sexually harassed, showing that, for example, 17.1% and 3.3% of women and men, respectively, had been verbally harassed since entering college/university. At a similar level, 15.2% and 4.4% of women and men, respectively, reported unwanted touching, hugging or kissing since they started on the college/university studies (see figure 1 for more details).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Detailed timeframe of sexual harassment forms stratified by sex. Error bars represent 95% CI.

As outlined in box 1 and further detailed in figure 2, all forms of sexual harassment experienced during the past year were significantly more common in the youngest age cohort. For example, while 19.4% of those aged 18–20 years reported any form of sexual harassment during the last 12 months, the corresponding proportion among those in the older cohort was 18.1% (21–22 years), 16.4% (23–25 years), 14.3% (26–28 years) and 9.3% (29–35 years). Similar age effects were observed across most forms of past-year sexual harassment, with a few exceptions (eg, indecent exposure).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Proportion of students across age groups reporting sexual harassment in the past year. Error bars represent 95% CI.

Number of occurrences of sexual harassment

As detailed in figure 3, the majority of those students reporting lifetime sexual harassment reported multiple occurrences across all forms of harassment, except rape, for which the majority of both women (68.2%) and men (66.1%) reporting rape had experienced this once. Among those reporting sexual verbal harassment, a larger proportion of women (36.1%) than men (22.7%) reported repeated verbal harassment (defined as more than 10 times). This gender difference was also present for intrusive eye or body movements, but not for the other harassment forms (see figure 3 for details).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Number of lifetime occurrences among students having reported harassment stratified by gender.

Person reported as perpetrators of the sexual harassment

The large majority of past-year sexual harassment acts (across all forms: 77%–86%) were conducted by someone outside of the university (neither a fellow student nor university staff member). Overall, fellow students committed the harassment in 18%–29% of the cases, whereas a university staff member was reported to have conducted the harassment in 0.6%–4.6% of the instances. As displayed in figure 4, among those students having reported each harassment form within the past year, a larger proportion of men compared with women reported being sexually harassed by a fellow student. A correspondingly larger proportion of women reported being sexually harassed by someone outside of the university, compared with men. There was also a trend that a larger proportion of men reported being harassed by a university staff member across most harassment forms (p<0.001), except for physical harassment.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Frequency of students, staff and others as perpetrators among male and female students who reported exposure to harassment in the past year. Sums may exceed 100% as categories are not mutually exclusive.

Statement of principal findings

This large national survey from 2018, inviting all full-time Norwegian university and college students aged 18–35, suggests that sexual harassment among Norwegian college and university students is prevalent. Almost one in four students (24.2%) reported lifetime sexual harassment of some form, while one in six students (16.7%) reported being sexually harassed during the past year. Women reported more lifetime and recent exposure to sexual harassment across all subtypes included in the survey, but also a substantial proportion of men reported exposure to sexual harassment. Younger students reported significantly more sexual harassment the past year than older students, and the majority of those being harassed reported multiple occurrences across all forms of harassment (except rape). The majority of recent harassment acts were committed by someone outside of the university setting, while a fellow student and university staff member committed the harassment in 18%–29% and 0.6%–4.6%, respectively, of the instances.

Strengths and weaknesses in relation to other studies, discussing important differences in results

Being one of the first studies of this magnitude on the prevalence of sexual harassment conducted after the #MeToo movement, the current study provides new evidence that sexual harassment indeed is a substantial problem also in college and university settings. The current estimates of sexual harassment are comparable with the review by Fedina et al,7 who synthesised research in the field from the last 15 years, concluding that although prevalence rates vary, they all indicate that a substantial number of college and university students experience sexual harassment. In line with that review, as well as estimates from European universities,14 15 we also found that unwanted sexual contact and verbal sexual harassment were the most prevalent forms, both being reported by approximately 15% of the students (20%–21% of women and 4%–5% of men). In terms of lifetime rape and rape attempt, this was reported by 3.4% and 2.1%, respectively, whereas the past year prevalences were around 0.4%. And while the results showed that most of the rape and rape attempts occurred before entering college/university, the majority of those students reporting the other five sexual harassment forms had experienced this after they started their college/university studies. The increased focus and attention following #MeToo may have impacted some of the respondents’ willingness to describe taboo experiences, including sexual harassment. As such, it is difficult to outline possible trends or compare with prevalence rates from previous studies in academic settings. Similarly, differences in the used operationalisations and sampling procedures make it difficult to directly compare our findings with earlier studies. Some of the discrepancies may also be attributed to cultural characteristics.

Young women were especially at high risk, a finding that is also in accordance with previous studies specifically addressing this group.21 22 Male victims of sexual harassment have been less studied, and although sexual harassment against women is a much more prevalent issue the present results are in accordance with previous studies that report less, but still a substantial rate of sexual harassment among male students.23 The reasons for not reporting sexual harassment or abuse may be the same for women and men (eg, intimidation, shame, fear of not being believed), but there may also be other reasons for previous under-reports of male victims of sexual harassment, such as cultural attitudes about manliness, masculinity and so on. It should also be noted that previous research has shown that women and men differ in the perceptions of sexual harassment, as men has a higher threshold for labelling an experience as sexually harassing.24 Regarding the observed age difference, this might partly be explained by recall bias, differences in perceptions of sexual harassment between older and younger students, and/or a healthy survivor effect where the (severely) harassed drop out of university in a young age. How much these mechanisms affect the prevalence is difficult to quantify, although differences in recall bias are likely to affect the past year prevalence to a lesser degree than lifetime.

A small percentage of the reported sexual harassments were from university staff members. Even so the current study shows that there are still many students who experience some form of sexual harassment from a person in power or in a trusted position. This is especially worrisome and may need to be specifically addressed in preventive programmes.

The increase in digital communication has been paralleled by an increased focus on sexual harassment through sending or viewing offensive pictures (eg, people sending ‘nude/dick pics’).25 Very few studies have investigated this form of sexual harassment, and the present results document that, although it is among the less frequent types of sexual harassment, it is still reported by 3.7% of the students. The rate was relatively higher ‘the past month’ and ‘the past year’, which may indicate that it is a growing problem. However, trend studies over time are needed to further shed light on this.

Compared with findings from work-life populations, where most studies include only harassment conducted in the workplace, the present findings are not dissimilar to those reported recently from a Danish union survey.26 This survey found that, when asked about specific situations of unwanted sexual attention, 26.5% of women and 25.7% of men had experienced some sort of unwanted sexual attention in the workplace during the past 12 months. However, most larger scale studies have included only crude overall single-item assessments of exposure to sexual harassment, which yield much lower prevalence estimates. In the union survey, only 2.1% of women and 0.3% of men reported being exposed to sexual harassment during the past 12 months when using such a single-item measure. Applying the sexual experiences questionnaire, a questionnaire which includes items regarding specific types of harassment similar to those included in the present study, Fitzgerald et al 27 reported prevalences of more than 70% for women and 30% for men from a US military sample in 1999. However, there is a need for more nationally representative studies using questionnaires assessing specific types of sexually harassing situations to further elucidate the magnitude of this problem within the workforce.

Meaning of the study: possible explanations and implications for clinicians and policymakers

The study has some important clinical implications, as there is substantial evidence showing that victims of sexual assault are more likely to suffer from short-term and long-term health problems.2 3 Our finding showing that the youngest female students are particularly at risk of being sexually harassed is disturbing. The higher risk of harassment among female students is especially problematic as previous research has shown that women display stronger reactions to sexual harassment than men.28 So although many universities and student welfare organisations run awareness campaigns that focus on preventing sexual harassment on campuses, the results of the present study suggest more should be done. Colleges and universities have a unique opportunity to reach many young adults. Future studies should include information on the situations where the harassment occurs to better tailor interventions. There is still limited empirical evidence for the effectiveness of universal prevention programmes,29 although there are interventions with promising results.30 31 Our finding that the youngest students were more at risk than older students suggests that the interventions should target new students specifically. The high prevalence of sexual harassment also has implications for victim services, which include providing both appropriate treatment and legal services. While prevalence rates and forms of sexual harassment may differ from campus to campus, prevention, intervention and victim service strategies require a detailed overview of the specific needs for each campus population.

Research from work-life settings has shown that while training of employees in the forms of antiharassment courses is important, additional measures are needed. To minimise harassment, organisations and institutions should nurture a common culture that rejects harassment and where all students and employees are equally entitled to a work environment without harassment. Supporting the effectiveness of an ‘anti-harassment’ culture, research findings show that a strong climate for conflict management in organisations is associated with lower prevalence of harassment, while also protecting against the effects of harassment when it occurs.32

Strengths and weaknesses of the study

The most important limitation of the present study is the relatively modest response rate of 31%, with little information about the characteristics of non-participants beyond age and gender distribution. Selective participation could bias the prevalences observed to the extent the selection was correlated with reports of sexual harassment. On one hand, it has been shown that non-participants of health surveys in general have poorer health than participants,33 and as those exposed to sexual harassment have more mental2 and somatic3 health problems the current results may represent an underestimation of the true prevalence in the target population. On the other hand, people are in general more prone to participate in a survey if the topic is relevant to them personally,34 and as the information material of the SHoT2018 study focused much on ‘how the students really are and feel’, one may speculate if this would lead to a higher participation rate of individuals who felt that the topic was of particular relevance to them.

As response rates are particularly important in prevalence studies, care should be taken when generalising the current findings to the whole student population. Rather, it may be more appropriate to emphasise the relative differences between men and women, as well as different age cohorts found in the current study, as these estimates are less prone to selection bias. Related to this is the 69% female composition of the sample, which may represent a bias for the overall estimates, as women are far more likely to report harassments compared with men. Also, this problem was mitigated by examining findings separately for women and men. It is possible that the use of a web-based survey approach contributed to the modest response rate, as electronic platforms typically yield lower overall participation rates when compared with traditional postal mail approaches, such as paper-based surveys or face-to-face interviews.35 Another limitation is that we did not use a validated instrument when assessing sexual harassment, which would have been preferable and make comparisons across studies easier. Rather, we designed a new assessment battery, carefully following the content of the Norwegian legal regulations, which also corresponds well with the formal scientific definition of sexual harassment.1 The main rationale for making a new instrument was due to recent recommendations in the review by Fedina et al,7 where the authors specifically called for studies specifying the timeframe in which the sexual harassment occurred and to provide brief examples of unwanted sexual experiences when defining sexual harassment. A final limitation is that we do not know the extent of missing data, as the questionnaire was designed to only indicate the presence of each harassment form, and not the absence. As such, although we consider it most likely that a missing answer is an indication of not having experienced that particular sexual harassment, missing answer may also mean that (1) the participant may not want to answer that particular item, (2) the participant does not remember it, or (3) the data may be missing at random, for example, not related to sexual harassment.

The most important study strengths include the very large and heterogeneous sample, as nearly all previous studies in this field focused on white, female, young undergraduate students.7 Also, notwithstanding being newly developed, the included sexual harassment instrument allowed us to explore more detailed information than previously assessed.