The analysis of Social Media data has moved from early excitement about their potential benefits to recognition that of the problems that they entail. As researchers, we need construct robust, reliable and valuable methodologies that build on their benefits while taking into account the complexities of working with these new forms of data. This workshop showcases an abductive approach to the research process – using social media data to explore, develop and refine research questions. Participants will be encouraged to bring along a “nano-research question” that can be explored during the workshop, using standard tools (Chrome Web Browser, Microsoft Excel) and other freely available open source software to collect and analyse Twitter data. The workshop will focus on four key aspects of the use of social media data in research: (1) Thinking Critically about the Social Data Pipeline (2) Acquiring and Examining Twitter Data (3) Understanding Twitter Content: Text Analysis (4) Understanding Twitter Social Engagement: Network Analysis.
Wasim Ahmed (Social Media Research Foundation, UK)
Introduction to Social Media Network Analysis with NodeXL
ABSTRACT. Prerequisites: Windows laptop with Excel
This short course provides an overview of social network analysis (SNA) via NodeXL and demonstrates through theory and practical case studies its application to research, particularly on social media and digital interaction and behaviour records. Wasim Ahmed, as a member of the Social Media Research Foundation and Connected Action Consulting, is an official representative of team NodeXL.
Design Exploration of Social Media Habits: Using information design research techniques to understand why people share on social media.
ABSTRACT. Building on the success of our Free Pile Sort Method Workshop at Social Media and Society 2017, this workshop will demonstrate multiple methods commonly used to assess the user experience of information and show how they can be applied to qualitatively understand why people share content online using the issue of fake news as a case study.
Scholars suggest that there are a number of indicators when content might be fake. For example, the University of Michigan (2017) identifies seven (7) types of mis- and disinformation. They are: Satire or parody, misleading information, imposter content, fabricated content, false connection, false context, and manipulated content. However, it is not clear whether everyday users of online content may be classifying information along these grounds when they decide what or what not to share online. We suggest that design is a crucial and underexplored part of how people consume different social media content, and thus seek to explore ways that researchers can use design principles to understand why people engage with social media content differently, sharing it in some cases, and in others, passing it by.
Our research on social media platforms indicates that the complex interrelationships of use, networks, and frequency influences perceptions of value (Traynor, Hodson and Wilkes, 2016), and we believe that equally complex relationships between users and information influence whether information is perceived as trustworthy. With this in mind, we will show participants several tools in the UX toolkit that could help researchers uncover the aesthetic and emotional motivators that drive people to share content. For this workshop, we will highlight how different methods are used for formative and summative testing and we will use the example of fake news as a case study to show how we would employ UX methods for relevant and insightful scholarly inquiry.
ABSTRACT. The workshop opens with a discussion of how to repurpose digital "methods of the medium" for social and cultural scholarly research, including its limitations, critiques and ethics. Subsequently participants are trained in using digital methods in hands-on sessions. How to use crawlers for dynamic URL sampling and issue network mapping? How to employ scrapers to create a bias or partisanship diagnostic instrument? We also consider how to deploy online platforms for social research. How to transform Wikipedia from an online encyclopedia to a device for cross-cultural memory studies? How to make use of social media so as to profile the preferences and tastes of politicians’ friends, and also locate most engaged with content? How to make use of Twitter analytics to debanalize tweets, and provide compelling accounts of events on the ground? Finally, the workshop turns to the question of employing web data and metrics as societal indices more generally.
Jacob Groshek (Boston University, United States) Li Zhang (Boston University, United States)
Building Your Own Cloud-Based Twitter Capture and Analysis Toolkit (TCAT)
ABSTRACT. ABSTRACT Objectives:
This tutorial aims to enable social media researchers to build their own cloud-based Twitter Capture and Analysis Toolkit, or TCAT (Borra & Rieder, 2014). The tutorial offers not only a detailed introduction to both the TCAT and the EC2 service of the AWS platform, but also a step-to-step guide to installing TCAT on their own EC2 virtual machine that will allow researchers to collect Twitter data at scale while paying only the cost for data storage. After participants have completed the installation, they will be working on projects of their choice using TCAT, while presenters will be available to offer help and suggestions.
Significance/Relevance: For researchers working with social media data, TCAT offers a host of powerful features to monitor, retrieve, process, and analyze Tweets in various ways. When run on a cloud-based virtual machine such as the EC2 service of AWS platform, it offers researchers additional flexibility and productivity.
Why a Tutorial/Workshop/Panel? To our knowledge, there is only limited tutorial on building a cloud-based TCAT available to the public. We feel there is a need for a detailed, step-to-step guide to walk through the whole process, especially for researchers with no or little programming experience. A tutorial will be an appropriate venue for this purpose.
Logistics: For the first 15 minutes, Dr. Groshek will introduce the TCAT interface and its main functions. After that, Li Zhang will be providing a step-to-step guide to building TCAT on the AWS platform. Both will then contribute to troubleshooting and offering insights into how to best utilize the data.
Prerequisites
If you plan to attend this workshop, please make sure to have the following:
Luke Sloan (Cardiff University School of Social Sciences, UK) Anabel Quan-Haase (University of Western Ontario, Canada) Dhiraj Murthy (The University of Texas at Austin, United States) Libby Bishop (GESIS, Germany) Grant Blank (University of Oxford, UK)
Linking Social Survey and Social Media Data: asking about it, doing it, and sharing it
ABSTRACT. The purpose of this session is to explore the opportunities available for linking social survey and social media data and what we need to consider when doing it. Sloan will open the workshop and discuss the potentials of linked data, exploring what insights an augmented dataset could give us into the social world drawing on a recent linkage experiment included on the Understanding Society Innovation Panel 2017. Quan-Haase will continue the discussion of linkage, focusing on issues of privacy and highlighting the limitations of what linked data might be able to tell us about specific subpopulations. She will draw on data collected from older adults in East York, Toronto, Canada to show the reluctance this population shows in joining social media and sharing on the platforms. She will highlight the relevance of privacy literacy for data sciences in moving forward with creating linked data sets. Murthy will provide an overview of the vast array of data available to us on Twitter, Facebook, and Instagram, demonstrating the type of insights that a linked dataset might be able to generate as well as the practical and ethical challenges this type of work entails. He will use samples of real datasets to enable participants to engage with these types of data hands-on. Bishop will then explore issues around the secure linkage of data, archiving and enabling sharing with the wider academic community.