- Anthropology & Sociology: The Social Worlds of Big Data
- Philosophy: Privacy, Objectivity and Norms
- Public Health: Health-related Behaviors in Social Networks
- Physics of Social Networks
Within the overarching framework of the Sensible DTU/Social Fabric research projects, a team of sociologists and anthropologists from UCPH will conduct a joint subproject, with the overall aim of pushing current boundaries for how to interrelate and cross-fertilize quantitatively and qualitatively based understandings of social networks. This will be done by exploring a several interrelated research questions, themes and methods at the core of current concerns in sociology and anthropology, as well as in the cross-disciplinary fields of computational, digital, and experimental social science.
Empirically, we seek to explain quantitative patterns of observed social practice, e.g. in terms of friendship formation and educational performance, by applying and further developing existing theories of social network patterns and effects. At the same time, we seek to enrich and challenge such quantitative and computational approaches by deploying ethnographic field work to the study of how friendship (and other social relations) emerge, and gradually develop and transform, among university student cohorts.
Based partly on these empirical questions and findings, we furthermore seek first answers to a number of profound methodological questions, of relevance to the future of the social sciences in an age of ‘big data’. Does the rise of computational social science, for instance, lead to a reconfiguration of the increasingly obsolete split between quantitative and qualitative research methods and data – and, if so, with what consequence for explanatory ambitions and models in sociology, anthropology, and beyond? What kinds of social scientific experiments does the Sensible DTU framework allow for, and how might such methodological innovations enrich existing scientific experimental designs?
Finally, we wish to include the Sensible DTU/Social Fabric research projects and the researchers partaking in them into our ambit of research by posing a variety of political and epistemological questions – influenced by Science and Technology Studies (STS) and related fields – concerning the rise of ‘big data’. What ethical, political and institutional challenges and opportunities does the rise of large-scale digital trace databases pose to the social sciences and society writ large? What may be learned about emerging dynamics of cross-disciplinary ‘big data’ research collaboration from the Social Fabric experiment itself?
The data collected in "Social Fabric" via smart-phones and surveys offer a unique opportunity to investigate human personality and social interactions.
The study of the relations between human social behaviour and personality constitutes a relatively new area within psychology. Thus, the present study aims to shed light on the associations between individuals' personality traits across time and context, and their everyday lives, behaviour, and social interactions.
The combined use of repeated psychometric measures and electronically collected “big data”, regarding individual behaviour and interactions over time, allows us to address questions such as “How does a person's personality and affective well-being influence his/her behaviour?”, “How do people use online social forums, and how does their online behaviour correlate with their offline behaviour?”, and “Which factors can account for the variability in our behaviour and personality across time, as well as in different contexts?”.
As part of the Social Fabric project, researchers from Section of Philosophy, Department of Media, Cognition and Communication, are carrying out three smaller projects. They concern the issue of privacy, objectivity, and norms respectively.
Research and the Right to Privacy
Research must be carried out in such a way that participants’ right to privacy is respected. It is because of this right to privacy that researchers typically make sure that all collected data are treated confidentially and anonymized such that participants’ identity is protected. Further, the right to privacy is taken into account in the process of collecting data. For instance, it is typically regarded as a violation of participants’ right to privacy if they are not aware that data about them is being collected; if they don’t grasp the extent of the information being collected about them; or if they feel watched over or invaded during the data collection process. This subproject has as its focus how it may be ensured that the right to privacy is respected during the collection of data: Under what circumstances is there a risk of violating the right to privacy when data about social networks is collected by way of a smart phone app and by use of participant observation? How may a violation of this right be avoided? And does it matter whether several methods, such as data collection by way of the smart phone app and participant observation, are employed?
In the social fabric project data about social networks are collected by way of both quantitative and qualitative methods: Data are gathered using the smart phone app (a quantitative method) and these data are supplemented by data gathered, among other things, by carrying out participant observation (a qualitative method). It is often held that so-called quantitative methods are more objective than qualitative ones. But is that is really so? The present subproject discusses this question while drawing on recent discussions of objectivity within the philosophy of science: What features are distinctive of an objective method? Is the collection of data by way of the smart phone app a more objective method than participant observation? To what extent is it possible to increase the objectivity of these two methods? And does it matter for the objectivity of both methods that they are combined as they are in the social fabric project?
Norms of interaction in social media
When interacting with others face-to-face, we have access to a host of social cues regarding other people and the social setting surrounding us: By means of our eyes, we can tell whether we find ourselves in a church room and, by the tone of voice of our interaction partners, we can hear if our actions aggravate them. In what way, then, is our ability to decode each others’ normative expectations affected when the interaction takes place online where our access to social cues is reduced to different extents? And, is there a behavioural difference when we interact with people inside our social network as opposed to complete strangers? Is there a connection between correct decoding of the normative expectations in a given situation, and whether or not we are personally acquainted with the others? Finally, in what respect can potential anonymity and geographical distance—both of which significantly reduce the risk of our actions having moral consequences—be said to affect our norm compliance? Those are some of the questions this subproject investigates.
Health-related behaviors like alcohol consumption, physical inactivity, smoking, drug abuse and disturbed sleep patterns represent major preventable risk factors for morbidity and mortality. In order to develop effective preventive programs, we need to know how such behavioral patterns are established and maintained.
There is a rising awareness that social relations may be central in this regard. For example research has suggested that social relations are associated with smoking habits, diet and compliance with medical treatment at the individual level. In general, healthier habits are more common among those with access to well-functioning social relations.
An interesting question, however, is if the characteristics of the social network as a whole influence health behaviors in a population. Highly discussed and cited studies have suggested that health-related behaviors spread through social networks like an infectious disease because we mimic the behavior of our peers. However, clustering of certain behaviors may also be due to the fact that we seek out peers who exhibit behaviors similar to our own.
The main challenge for all studies of behavior is to disentangle the contagious effects (i.e. behavior spreading from one individual to the next through the network) from the clustering of behavior that is due to individuals seeking out relations to others with similar behavior (sometimes called latent homophily). Previous studies in this area have been criticized for not being able to separate these complex feedback mechanisms between social networks and health-related behavior.
Through The Social Fabric/Sensible DTU project we aim to establish how health-related behavior spreads through or emerge in newly formed social networks of 1000 freshman students repeatedly followed for changes in both health-related behavior and their social networks over a full academic year. Focusing on social networks as they develop among new university students considerably adds to the existing literature, which has almost exclusively looked at existing networks. Addressing newly formed networks will enable us to distinguish between contagious and homophilic effects. Moreover, focusing specifically on a young group of freshman students will provide a unique opportunity to follow early establishments of behavioral patterns that may persist many years into adult life and potentially have a major impact on the burden of disease.
Combined, these features provide an unprecedented possibility to gain insights into the mechanisms that determine our behavioral patterns in early adult life, which is essential for guiding future preventive strategies.
A team of economists participate in the framework of Social Fabric to explore, in cooperation with other groups, the role of social networks in determining peer influence and peer selection. Understanding Sensible DTU provides a novel way to measure social interactions on a large scale, which constitutes a unique data platform to further this research agenda.
The project seeks to examine the process of friendship and study groups formation among university students. A specific aim is to quantitatively identify how the process of social tie formation relates to students’ individual characteristics and how these may lead to different kinds of ties, which have different degrees of stability over time. Another aim is to explore the role of communication in social ties and groups on and how this quantitatively affect study behavior, academic outcomes etc. As a separate goal, we hope to further our understanding of study behavior and drop out behavior in tertiary education.
A team of physicists from the Niels Bohr Institute will employ methods from statistical physics and information science to analyse the data generated in the Sensible DTU experiment. They will investigate the nature of social networks and how the form and develop over time, and wish to quantize the influence of individuals upon each other, as well as the possible spread of behaviour across different information channels.