Abstracts

From ‘big data’ to hybrid digital methods: Digital media and communication research in the post-API era

In this age of ‘deep mediatization’ (Couldry and Hepp, 2017), a handful of proprietary digital media platforms owned by massive technology companies play a central role in shaping and contesting society and culture. Over the past decade, researchers from a range of HASS disciplines have developed new computational and hybrid digital methods to map and interrogate these developments. Much digital methods work relied on gaining access to data through platform APIs (Application Programming Interfaces). These tools were generally created by proprietary platforms as part of the mid-2000s Web 2.0 paradigm and its associated ethos of open innovation. But the innovation ecosystem is rapidly evaporating and API access is progressively being shut down, in the wake of a series of scandals around platform privacy, extremism and hate speech. Meanwhile, publics are increasingly communicating via visual, ephemeral, and private forms and platforms; at global, local, and intimate scales. In this talk, I map out some of the possible futures of digital methods in this environment, including new hybrid approaches to platform analysis and computational critique; and secondly, by suggesting some of the parameters for research ethics, which will require institutional coordination and support.

The App Walkthrough – Jean Burgess

Software applications (apps) are the site of significant sociocultural and economic transformations across many domains, from health and relationships to entertainment and finance. As relatively closed systems, apps pose methodological challenges for digital media research. In this session, we will discuss the walkthrough method approach, which combines cultural studies and science and technology studies (STS) as a lens for critical app analysis. Participants will learn how to establish an app’s environment of expected use by assessing its vision, operating model, and modes of governance. They will also gain hands-on experience using the walkthrough technique to systematically step through the stages of registration, everyday use, and discontinuation to identify the app’s embedded cultural meanings and implied ideal users, as well as to identify traces of its data flows and algorithmic logics.

The word Izolo means yesterday in the South African language isiXhosa. Our method is about creating a detailed picture of what yesterday’s mobile phone practices looked like for one particular person through a long interview.  In this interactive workshop, participants will learn how we used this method to understand the mobile practices of the “less-connected” in South Africa, marginalised users of the mobile internet who primarily use data-light applications like WhatsApp. We will also explore some of the ethical and practical challenges we encountered.

Diary methods, while widely used in media studies over many decades have proved difficult to adapt for studying mobile phone use. The method we developed uses a reconstructed diary approach. It involves a guided interview where researchers aim to elicit a detailed description of the everyday practices and contexts of mobile communications during one day or period of the day. The responses are captured both in note format on the questionnaire and as recorded interviews to provide more detailed qualitative data. We will focus on sections of the questionnaire that deal with capturing holistic scenes of daily life and the mobile practices that form part of such scenes, mobile practices that may be both functional and symbolic.

You will have a chance to work with an original questionnaire from our project to try and capture a picture of everyday mobile use of your colleagues in the workshop. Participants will report back on their experience of using the method. The session will end with a discussion around how one analyses the data to build holistic understandings of mobile practices among the “less-connected” and whether this method can be usefully used or adapted for other studies and  contexts.

* The Izolo Diary Method was developed by Indra de Lanerolle, Marion Walton and Alette Schoon.

Abstracts:

Are Filter Bubbles Real?

The success of political movements that appear to be immune to any factual evidence that contradicts their claims – from the Brexiteers to the ‘alt-right’, neo-fascist groups supporting Donald Trump – has reinvigorated claims that social media spaces constitute so-called ‘filter bubbles’ or ‘echo chambers’. But while such claims may appear intuitively true to politicians and journalists – who have themselves been accused of living in filter bubbles –, the evidence that ordinary users experience their everyday social media environments as uniform and homophilous spaces is far more limited. For instance, a 2016 Pew Center study has shown that only 23% of U.S. users on Facebook and 17% on Twitter now say with confidence that most of their contacts’ views are similar to their own. 20% have changed their minds about a political or social issue because of interactions on social media. Similarly, large-scale studies of follower and interaction networks on social media show that such networks are often thoroughly interconnected and facilitate the flow of information across boundaries of personal ideology and interest, except for a few especially hardcore partisan communities. This talk explores the evidence for and against echo chambers and filter bubbles. It moves the present debate beyond a merely anecdotal footing, and offers a more reliable assessment of this purported threat.

Social Media Analytics: From Raw Data to Engagement Metrics

This data analytics and visualisation workshop introduces a number of standard tools and methods for large-scale data analytics, using Twitter data to illustrate these approaches. The workshop introduces participants to the open-source Twitter Capture and Analysis Toolkit (TCAT) as a capable and reliable tool for data gathering from the Twitter API, and to the high-end data analytics software Tableau as a powerful means of processing and visualising large datasets. The skills gained in the workshop are also transferrable to working with other large datasets from social media and other sources. The workshop is suitable for participants new to working with social media datasets, and as a refresher for more experienced users.

Abstract

Computational Power: Automated Use of WhatsApp in the Elections

The 2018 Presidential Elections has raised great concern in Brazilian media, academia and society regarding the strategic use of political campaigning on social networks, mainly the messaging application WhatsApp. In this context, this study investigates the primary factors that demonstrate the degree of coordination among WhatsApp groups and identifies patterns of behavior of users who are disseminating content. To this end, we monitored 110 open political groups on WhatsApp for a period of one week, and we analyzed the list of members and messages to answer three questions: 1) if there are indications of automation being used to send messages; 2) if there are indications of coordinated distribution of information among WhatsApp groups; and 3) what the degree of connection among participants and administrators of the political groups on the platform is. Our study concluded that there is strong evidence of automation being used in multiple WhatsApp groups and that there is a high degree of interconnection, as evidenced by the large number of administrators and members shared by these groups among each other.

Abstract

The poverty of methods in studying ‘the internet of the poor’

In this talk, I consider how the ‘internet of the poor’ – the internet that most people in South Africa and Africa experience – poses new and important questions to internet researchers. Do we have the right tools to study this phenomena? And how do we need to apply them?

This feels to me to be an urgent and unanswered question. Internet time and academic time are, unsurprisingly perhaps, out of sync. The major internet giants are changing the objects of study more quickly than we develop approaches to researching them. Our research approaches and access are  constrained by those who control the major digital platforms. The resources of university researchers are dwarfed by the research capabilities of those corporations that own the networks we study.

These are global problems, but in Africa there are particular issues of method that demand responses. As Jonathan Donner observes, just as the internet comes within reach of the majority of people on the planet, the internet they participate in and which we study may not be the same internet. This ‘other’ internet we have described as the internet of the poor. Studying this other internet poses methodological challenges. I examine three of these challenges drawing on examples of internet research in South Africa and other African countries over the last decade.

The first challenge is that while we need reliable quantitative survey research to understand how South Africans and Africans use, access and shape an African internet, we struggle to produce such data. The cost and complexity of selecting representative samples is much greater in Africa than in many other parts of the world. Data inequalities precede and are being reproduced in digital inequalities.

The second challenge concerns observational bias – we are tending to look in places that are easier to see. The Twitter API gives us access to data that enables us to use research methods that we cannot apply to, for example, WhatsApp which is far more widely used in South Africa and on the rest of the continent.

The third challenge is that researchers on the continent are struggling with the methodological innovations required to explore and understand the phenomena we are examining. Our capacity – and our desire to learn about – statistical and network analysis tools for example, is limiting our research capabilities.  And just as we are recipients of global internet phenomena we are also recipients of global research methodologies. We are not using our understanding of our own context to select or develop appropriate methods. Where we do, we are failing to share these amongst researchers on the continent and beyond.

This Symposium comes at a crucial time. It offers opportunities to develop our technical knowledge and skills. It also offers the opportunity to set an agenda for collaborating in developing a new toolkit of research methods appropriate for understanding the internet of the poor.

Abstract:

Social media apps and everyday life in South Africa

Over the past decade, social media platforms have deeply penetrated the fabric of everyday life. The growth of the internet has led to the proliferation of social networking sites such as Facebook and Twitter; and despite ongoing issues related to the ‘digital divide’, increased access to the mobile internet in Africa has resulted in more people using their mobile phones to go online; and the consequent growth and popularity of social media. In this talk I  explore how social media is used in South Africa, through briefly touching on a range of platform specific case studies exploring e.g. Strava within the context of the sociocultural practice of contemporary self-tracking practices in the African context, Tinder and the visual politics of online dating and the digital transformation of intimacy; as well as exploring

location based apps such as Instagram and how these result in geocoded digital traces which can be analysed to access the complexities of the urban experience.

More specifically, I consider social media as a reflection of people’s private lives, and not just a source or expression of political power; keeping in mind that social mediation transcends the personal and political, so that sometimes politics emerges from the personal. In this talk I reflect on social media and everyday life in South Africa, doing digital research in South Africa, and drawing on concepts such as ‘mundane data’ (Lupton, 2016) to explore how digital data plays a central role in purposing and repurposing concepts of the self, how social relationships are configured and negotiated via data; and how people incorporate data into their practices and concepts of selfhood and embodiment; while simultaneously foregrounding the “agency and reflexivity of individual users as well as the variable ways in which power and participation are constructed and enacted” (Pink et al 2017, 2).

Mecodify Demo

This workshop will introduce participants to Mecodify, a bespoke Twitter analysis tool freely available on GitHub. Using Mecodify’s web search method for data collection produces the same results as when using the web search through Twitter’s Advanced Search page. Mecodify crawls the Twitter search page via a built-in script and extracts the tweets pulled up by search queries. It then fetches all messages and relevant information about the users who tweeted via the Twitter Application Program Interface (API). Mecodify can be used in conjunction with the open source tool Kumu, for social network analysis.

Zimbabwe like most African countries is a multicultural nation with multiple ethnolinguistic groups – groups that are unified by both a common ethnicity and language. The country is characterised by a complex language ecology which comprises multiple ethnolinguistic identities and the politics of language and belonging are contentious issues.

The subject of ethnicity is sensitive and is to some extent associated with undertones of regionalism and division. In the Zimbabwe context any attempt to discuss ethnicity risks being labelled as tribalism and, therefore, divisive to a ‘supposedly’ united nation. It is against this background that disenfranchised linguistic minorities have taken advantage of the liberative potential of social media. They are regrouping in the virtual space and in the process forming vibrant ethnolinguistic online communities such as #NdebeleTwitter on Twitter. This research seeks to critically analyse the motivations and purpose of this particular Zimbabwean ethnolinguistic online community. This is done by critically analysing #NdebeleTwitter by examining the online community by posing questions such as;

1. Who are the prominent users, and what are they frequently commenting on when using the hashtag #NdebeleTwitter?
2. What types of interactions emerge under the use of the hashtag?

iii.            What interpretation patterns can be distinguished in the way users engage with one another on the hashtag?

1. What types of prominent themes can be identified?
2. What are the determinants that an individual is part of #NdebeleTwitter?
3. How do users under the hashtag engage with other users outside of #NdebeleTwitter, in particular those who hold a critical perspective?

The research will use the application tools Mecodify and Twitonomy examine information flow, influence, opinions and sentiments. Mecodify is an open-source tool for simplifying big data analysis and visualization, and Twitonomy is an analytics tool that provides insights and features that help monitor, manage, track and optimize activities on Twitter.

Digital data has afforded researchers novel modalities of researching almost all fields of study. Researching using digital data, just like data that involve people, requires digital researchers to be aware of the complex power relations, ethical issues, voice and accountability. Methodological context, on the other hand, plays an integral part in generation/collection and interpretation of data. This paper contains my methodological reflections of researching in and about Kenya for the last seven years, and suggest essential lessons for digital researchers. It is important to note that Kenya, as a postcolonial nation carries important historical past of pre and postcolonial. Also, socio-political and cultural issues become embedded in digital spaces and thus affect how certain digital data is generated and analysed. For instance, ethnic and religious diversity, political affiliations, class dynamics and gender issues are vital issues that should be considered by digital researchers researching about Kenya. Kenya has an ethnic and religious plurality that is politicised since they become epicentres of power contestations since independence at a regional and national level. Interestingly, these issues do not affect all geographical locations in Kenya uniformly since most of these regions do not share the same socio-political and cultural issues. Having an understanding of these contextual issues became and have continued to be important in my research because they shape my methodological approach as well as the interpretation of research findings. Understanding these contextual issues in digital media spaces has the implication of increasing digital media research output in Kenya and about Kenya since, like many other nations in sub-Saharan Africa, research in digital media technologies remains thin.

Key Words: Digital Media, Methodology, Kenya, Context

The Abakuria speaking people of Kenya are numerically one of the minority communities constituting just under 0.58% of the country’s population. This numerical disadvantage has meant that their voices have been relegated to the margins of socio-cultural and political discourse in Kenya. As of 1 January 2016, the Abakuria of Kenya did not have a single radio FM or television station broadcasting in their language. This, in contrast to certain dominant language groups that have had several broadcasting platforms in their languages and dialects for over 20 years now. This study, therefore, sought to determine the extent to which ordinary citizens – especially those on the margins of the dominant centre – could actively participate in the (co)generation, (co)production and (co)distribution of broadcast content, and to determine what value such participation could offer them, both culturally and socioeconomically. To achieve this objective, several possible approaches and methodologies through which the Abakuria community could actively access a platform for self-expression and self-representation in the digital media scape were considered resulting into an adapted confluence between Participatory Action Research and Digital Storytelling approaches. The pilot project entailed working with eight “ordinary members” of the community over eight days in a process that resulted in the production of short films and photography for digital broadcasting. The participants received technical training workshops on conceptualising and crafting their narratives; basics of photography, cinematography, video and image editing; other basic postproduction skills such as titling and subtitling; and an introduction to some digital platforms that could be considered for broadcasting.

Key words: Digital storytelling, self-expression, self-representation, Active User Generated

Viewers of South African reality dating shows have found a platform through Twitter to radically change the manner in which they consume content of the shows. Audiences of Date My Family and Uyang’thanda Na use this social media platform as an opportunity to express their views and judgements about the show and in particular about the performances of those seeking love on the show. The traditionally private act of courtship in an African context is reimagined in the manner in which audiences in these shows have brought the scrutiny of private act into the public realm. Through a frame analysis of one of the first noteworthy episodes of Date My Family and subsequent value judgement by viewers on Twitter, this paper argues that there is nothing inherently surprising about the seeming shift of power to the audience in authenticating contestant performance on South African reality dating shows.

Identity Creation

“Digiphrenia,” which is characterised as the experience of trying to exist in more than one incarnation of yourself at the same time is an unintended after effect of the presentation of one’s ideal self, online. These fragmented identities often manifest as warped representations of the real self. Online movements such as #InstagramVSReality have however subverted this narrative. A desktop content analysis on a sample group of these images is indicative of how Instagram users are increasing self-awareness in terms of their online identity creation and are striving for authenticity in an online world.

Index Terms—Instagram, Online Identities, Digiphrenia, Social Media, Online Movements, Reclaiming Authenticity

Ethnographic methods for researching the shifts in South African journalism in an age of digital interactivity

In this interactive workshop veteran South African journalists Heather Cameron and Gus Silber work with their supervisor Prof Harry Dugmore to present participants with various scenarios from their qualitative research. The workshop involves a range of discussions on how to build theory from rich qualitative data. Using examples from Gus’s research, the participants will share the intimate world of news consumption and digital circulation among  various Johannesburg families. Heather’s research explores the process of creating alternative expert content by South African legal scholars whose blogs are widely consulted by both journalists and the general public for complex legal analysis of everyday politics. Linking their data to theories from digital news production and distribution, allows us to interrogate established ideas around networking and news, public spheres and private life, prosumer news media channels and the conceptual borders between amateur and professional news production. Most of the focus on methods tends to focus on data gathering. By focusing on the process of data analysis, we hope to stimulate debate among participants in terms of how they analyse their own data.