“Beyond Interaction Networks: An Introduction to Practice Mapping”
Axel Bruns;
Plenary
Social network analysis has become a key tool in digital communication research, and has flourished especially in social media studies using Twitter data. However, it struggles in analysing activity patterns on platforms that provide fewer data points on interactions between accounts (e.g. Facebook or Instagram), and in exploring the interconnections between multiple activity practices that are interwoven with each other. This keynote introduces the new approach of practice mapping, which advances beyond the network analysis and visualisation of direct interactions between accounts and instead uses vector embeddings of networked actions and interactions to map the commonalities and disjunctures in the practices of social media users. In particular, this innovative methodological framework has the potential to incorporate multiple distinct modes of activity and interactivity into a single practice map, can be further enriched with account-level attributes such as information gleaned from textual analysis, profile information, available demographic details, and other features, and can be applied even to a cross-platform analysis of communicative patterns and practices. Drawing on a case study of public posting activity on Facebook during the Voice to Parliament referendum campaign, this keynote outlines the practice mapping approach and demonstrates the insights it can produce.
“Indigenous Data, Indigenous Methodologies and Indigenous Data Sovereignty”
Maggie Walter;
Plenary
The field of Indigenous methodologies has grown strongly since Tuhiwai Smith’s 1999 groundbreaking Decolonizing Indigenous Methodologies. Contributions to scholarship and practice have emerged from First Peoples’ researchers across the globe including my own co-written (with Chris Anderson) 2013 book Indigenous Statistics: A Quantitative Methodology (Routledge). In this presentation I build on the two central premises of that writing. First, statistics are culturally embedded phenomena not neutral data. Across the Anglo colonized world, Indigenous data reflect the purposes and assumptions of those who commission, analyse and interpret them, not their subject. The result is a trope of deficit rather than the robust complexity of Indigenous lived realities. Second, methodology, rather than statistical methods, create these (non-Indigenous) culturally loaded data. The common failure to name up the methodology, as opposed to the method, used, normalises deficit data outcomes. I then use Indigenous Lifeworlds theory and the foundational principles of Indigenous Data Sovereignty to make the case for a fundamental disturbance of Western data logics via a mandatory recognition of methodology in Indigenous quantitative research.
“Mixed Methods: Finding a Place for Qualitative Research”
David Silverman;
Plenary
PUBLIC LECTURE [FOLLOWED BY Q&A]
There is increasing interest in collaborative research engaging different disciplines/sectors and methods. Are there some proven approaches in planning a large collaborative project involving both qualitative and quantitative research that ensure the qualitative contribute is optimized?
I begin by describing the appeal of mixed methods. Using examples from both academic journals and student research projects, I discuss studies that mix both qual and quant or just mix different qualitative methods.
Most mixed methods research begins with quantitative data and then moves on to qualitative materials, usually open-ended interviews. I criticize the assumption that this allows us to ‘go deeper’ or more empathetically into social phenomena.
I then outline the limitations of the suggestion that, by using mixed methods, we can reveal the ‘whole picture’ of some social phenomenon. I also draw attention to the naïve positivist assumptions behind much mixed methods research.
Quantitative research must define its variables at the outset in order to measure them reliably. In my view, the beauty of qualitative research is that, mostly using naturalistic data, it allows us to understand how social phenomena are put together rather than to legislate their character at the outset. I conclude by demonstrating a more fruitful division of labour between quant and qual research through which we can gain by mixing methods.