When is a Like a Like? Data-Generating Processes in Online Communication

Research question I: What different qualities, contexts of origin and processes underlie the relationship structures of online communication?

Research question II: How can the data-generating processes on online platforms be theoretically and empirically differentiated, described and explained?

Online communication offers a variety of ways for users to connect with one another. Hyperlinks have been a constituent component of the World Wide Web since its emergence, while on social networking sites, friend or contact lists come into play. In addition, liking, commenting and sharing messages mark relations among actors. These markings are, on one hand, functionally inscribed in online communication platforms. On the other hand, they result from users’ behaviour and subsequently orient other users’ behaviour. This project studies these markings from three perspectives.

First, the practice of marking relations is analysed to reconstruct and systematise the variety of markings. So far, it has been generally unclear how formally identifiable relations (e.g. hyperlinks, contacts and likes) can be interpreted. This lack of clarity becomes obvious in cases in which content such as journalistic reports on violence on Facebook is marked with ‘likes’ without it being possible to assume that every user has a positive attitude towards the violence. After content analyses of messages containing relations qualitative interviews will be conducted to gain insights into the underlying communication processes. In addition to analysing relationship markers from the perspective of the discloser or the recipient, comparing different platforms brings the specifics of the mediation contexts into view.

Second, a methodological problem is addressed. Marking relations results in standardised metrics and data structures, so they can be processed efficiently in an automated manner, for instance, to collect hyperlink networks in research projects. It is commonly assumed that technical links indicate substantive relationships between actors. Through a combination of automated and interpretative methods, this project questions the conditions under which valid statements about relationships can be made. Network data collected by automated methods are examined with interpretive methods to investigate to what extent metrics are still linked to users’ underlying behaviour or constitute a reality on their own.

Thirdly, the temporal processes and action sequences are examined in order to identify, differentiate, describe and explain typical and atypical processes on online platforms. At the action level, we ask which communication processes and sequences on online platforms are typically to be expected and which are unusual. From this, an inventory of temporal phenomena such as 'shitstorms', 'candy storms' or viral processes is created. A comparison of seven social media platforms will be used to explore the extent to which the different organizational methods and architectures of online platforms are reflected in the visible temporal processes. To this end, methods will be developed to visualize and analyze communication processes and sequences on online platforms.

Overview
© LIKE-Team

This project thus provides a communication sociological analysis of relationships in the online world and an epistemological evaluation of the scientific inquiry process. These analyses are needed to understand the formation of social relationships in socio-technical contexts and to establish automated procedures in the social sciences. Overall, the project reflects on a society increasingly reliant on deriving evaluations from data-generating processes.