The aim of this study was to investigate and describe online interpersonal victimization (OIPV) in terms of gender differences and the association between such victimization and online behavior of active social media users in a Swedish sample. Since social media has become such a big part of our world it is of importance to study OIPV in this forum. Previous research has found that OIPV is a rather common phenomenon, that there are gender differences included and that certain online behaviors are risk factors. OIPV by itself is not a crime but rather an umbrella term including the legal terms illegal threat, slander, insult, harassment, sexual harassment, stalking and crimes against the personal data act or the copyright act. The cyberlifestyle–routine activities theory was used in this study to understand which online behaviors were risk factors in our sample. To answer the aim a survey was made and answered by 338 participants. The answers were tested with chi-square tests (χ²) and Mann-Whitney U tests in order to examine differences in gender regarding victimization and to find differences between the victimized and non-victimized group regarding their online behaviors. The results showed a high prevalence of OIPV and that women were more likely to be victims of OIPV, especially of harassment, sexual harassment, threats of sexual violence and stalking. The online behaviors that were significant risk factors in our sample were the use of a profile picture of themselves and number of hours spent on social media every day. This combined indicated that social media may not be a completely gender equal place and that online behaviors may not indicate the risk of being victimized equally well for both genders. / <p>2016-06-01</p>

Essay #1: Reconstructing Online Behavior through Effort Minimization Data from online interactions increasingly informs our understanding of fundamental patterns of human behavior as well as commercial and social enterprises. However, this data is often limited to traces of users' interactions with digital objects (e.g. votes, likes, shares) and does not include potentially relevant data on what people actually observe online. Estimating what users see could therefore enhance understanding and prediction in a variety of problems. We propose a method to reconstruct online behavior based on data available in many practical settings. The method infers a user's most likely browsing trajectory assuming that people minimize effort exertion in online browsing. We apply this method to data from a social news website to distinguish between items not observed by a user and those observed but not liked. This distinction allows us to obtain significant improvements in prediction and inference in comparison with multiple alternatives across a collaborative filtering and a regression validation problem. Essay #2: Measuring Individual differences: A Big Data Approach The amount of behavioral and attitudinal data we generate every day has grown significantly. This era of Big Data has enormous potential to help psychologists and social scientists understand human behavior. Online interactions may not always signify a deep illustration of individuals' beliefs, yet large-scale data on individuals interacting with a variety of contents on specific topics can approximate individuals' attitudes toward those topics. We propose a novel automated method to measure individuals' attitudes empirically and implicitly using their digital footprints on social media platforms. The method evaluates content orientation and individuals' attitudes on dimensions (i.e. subjects) to explain individual-content ratings in social media, optimizing a pre-defined cost function. By applying this method to data from a social news website, we observed a significant test-retest correlation and substantial agreement in inter-rater reliability testing. Essay #3: Social Media and User Activity: An Opinion-Based Study An increasing fraction of social communications is conducted online, where physical constraints no longer structure interactions. This has significantly widened the circle of people with whom one can interact and has increased exposure to diverse opinions. Yet individuals may act and respond differently when faced with opinions far removed from their own, and in an online community such actions could activate important mechanisms in the system that form the future of the outlet. Studying such mechanisms can help us understand the social behaviors of communities in general and individuals in particular. It can also assist social media outlets with their platform design. We propose models that capture the changes in individuals' activities in social media caused by interacting with a variety of opinions. Estimating the parameters of the models using data available from a social news website (Balatarin) as a case study, we extracted mechanisms affecting the communities on this platform. We studied the effect of these mechanisms on the future formation and the lifecycle of the platform using an agent-based simulation model. Having examined the effect of biased communities on the social media, the results imply that individuals increase their online activity as a result of interacting with contents closely aligned to their own opinion. / Ph. D.

No description available.

No description available.