Matti Näsi, University of Turku, Finland
Atte Oksanen, The National Research Institute of Legal Policy
Emma Holkeri, The National Research Institute of Legal Policy
In the article, we intend to examine how being a victim of online harassment associates with the likelihood of a person accessing harmful online content, especially content relating with self-harm and eating disorders. We will also examine the same association with respondents’ perceived subjective wellbeing (SWB), social media activity, offline social activity, as well as age and gender. Furthermore, we will construct a cross-national comparison by comparing data from two countries, US and Finland. In this way, it is possible to establish a more comprehensive understanding of how such negative online behavior is present in different cultural contexts. As a result, the following research questions were formulated:
RQ1: What type of associations exists between being a victim of online harassment and visiting different types of harmful websites regarding negative self-image and self- harm in both the US and Finland?
RQ2: Is there a statistically significant association with online harassment after controlling for the other independent variables in the analysis in both US and Finland?
RQ3: Are there differences of association between US and Finland in terms of the different independent variables?
Data and Measures
Our data was collected from two countries, US and Finland, among participants aged between 15 and 30 years in the spring 2013 (US n=1032, Finland n=555). The participants were selected/recruited from a panel of Americans and Finns who participate voluntarily in different research surveys. The data collection was administered by Survey Sample International (SSI). The sample quota was calculated to be nationally representative on age, gender, and education for both countries and, in the case of the US, also in terms of race and appropriate regional coverage.
In terms of the dependent measure regarding whether the respondents were victims of online harassment, we asked the following question: “In your own opinion, have you been a target of harassment online, for example when people have spread private or groundless information about you or shared pictures of you without your permission?” with either yes or no response. Our central independent variables were questions concerning exposure to certain harmful online websites, particularly in relation to self-harm or negative self-image. This was measured by asking respondents whether they had seen any of the following types of websites: “Sites about ways of physically harming or hurting yourself”, “Sites about ways of committing suicide”, “Sites about ways of being very thin (e.g. sites relating to eating disorders)”, with either yes or no response. We also measured respondents’ age, gender, perceived subjective wellbeing (SWB), social media activity, and offline social activity.
Analysis
The statistical analysis included basic descriptive analysis which was conducted by applying cross-tabulation analysis. Furthermore, we also conducted logistic regression analysis. The descriptive analysis provided statistical information concerning the association whether respondents were victims online harassment or not, and how this was reflected in whether they had witnessed harmful online content, respondent social media activity, perceived levels of subjective wellbeing, offline social activity, along with age and gender. For the explanatory analysis, we used logistical regression to first examine the association between online harassment and having viewed self-harm or negative self-image themed websites for both the US and Finland. We then controlled for the rest of the independent variables.
Results
According to the descriptive statistics of the association with online harassment and having viewed self-harm or negative self-image websites, in both the US and in Finland, those who had been victims of online harassment had more likely visited such sites than those who had not been harassed. In regard to the other independent variables, of those harassed online, in the US, males were slightly more likely victims compared with females, whereas in Finland, females were more likely victims than males. In the US, the youngest age groups were more likely online harassment victims, whereas in Finland, the youngest and oldest respondents were the more likely victims. Those more active users of different social media were more likely victims in both US and Finland compared with those less active. In the case of subjective wellbeing, again in both US and Finland the victims of online harassment reported lower subjective wellbeing compared with those who were not victims. Frequency of offline social activity in the US appeared relatively invisible, whereas in Finland, those socially less active were more likely online harassment victims compared with those more socially active.
Further analysis showed that the association between the victims of online harassment and respondents who had witnessed sites about ways of physically harming or hurting themselves were significant in the US data compared with those who had not visited such sites. Furthermore, respondents who had visited sites about ways of being very thin were significant in both US and Finnish data compared with those who had not witnessed such websites. After controlling for the other independent variables, these same associations persisted. In terms of the other independent variables and their significance, in the US, age and SWB, and in Finland, gender and SWB were statistically significant.