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

• Applicant: Professor Dr. Jakob Jünger
• Subject Area: Communication Sciences
• Term: since 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 417844307

In the LIKE project, we are investigating the meanings behind reactions such as likes, forwards, and comments on online platforms. We start from the observation that these data points are used as metrics to quantify social behavior – in scientific studies, for example, when investigating political communication, but also in the non-scientific world. They document behavior, but at the same time, they affect activities when used by platforms to rank and display content, or when users take them as cues. However, aggregating and abstracting behavior to highly standardized indicators, particularly when they are collected and analyzed automatically, obscures the diversity of action orientations – for example, a like may indicate agreement with a statement in terms of content, express appreciation for a person, or simply result from routines when scrolling through incoming notifications. We investigate the different meanings with content analyses and interviews for the platforms Disqus, Facebook, Instagram, Twitter, and YouTube. So far, we have focused on the meanings of these data traces and thus on the origins and results of data-generating processes. Following on from this, we will now examine the processes themselves on three levels: 1. Action level: which temporal communication patterns and sequences are typically expected on online platforms, and which are exceptional? 2. Mediation level: to what extent are the different organizational modes and architectures of online platforms reflected in the visible temporal processes? 3. Data level: which methods can be used to reveal communication processes and sequences on online platforms, especially to distinguish typical from unusual phenomena? As a result, an inventory of temporal patterns and action sequences (e.g., trends, waves, spikes, boosts, take-offs) will be compiled. The inventory will allow us to assess how processes differ between different platforms and to what extent they reflect user behavior, platform decisions, or practices of scientific data analysis. By describing and explaining both the typical and exceptional processes on online platforms, the phenomena are substantively captured to develop a benchmark and methodological approaches for subsequent online studies.

• DFG Programme: Research Grants