Seeing Data: Are Good Big Data Visualisations Possible? January 2014 – September 2015

Overview

This collaborative research project aims to explore what factors impact the effectiveness of big data visualisations, using migration as a key topic. Big data are increasingly ubiquitous and are assumed to have the power to explain our social world. But despite our increasing exposure to dig data visualisations, very little is known about how they are received, whether they are effective, whether effectiveness is possible, and, if so, how effectiveness might be understood in this context. Little is also understood about the skills and literacies people need in order to make sense of them.

The project’s empirical research takes two different sets of migration data used by COMPAS’s Migration Observatory as a case study. Using both numerical 2011 Census data about the UK’s foreign-born population and textual data about newspaper portrayals of migrant groups, the team recruited field-leading data visualisers Clever Franke to design and produce visualisations of MigObs data. Then, it examined the reception of these visualisations among a range of other publicly available visualisations through in-depth focus group discussions with members of the general public and diaries documenting their continued engagement with visualisations seen elsewhere. It also examined how visualisation professionals impact the process of creating these objects, as well as defining what ‘effectiveness’ might actually mean.

Findings from the research will help MigObs address some of the challenges it faces in clearly communicating its data to a range of stakeholders, as well as contribute to key debates in science and technology studies about how visualisations are received.

Topics

Public Opinion

Regions

Europe

Methods

Seeing Data enlisted the help of Clever Franke, a design agency based in the Netherlands, to produce two visualisations based on data used by the Migration Observatory: characteristics of the UK’s foreign-born population according to the 2011 Census, and newspaper portrayals of migrant groups in the British press.

These two datasets were chosen for several reasons:
• They were sufficiently complex and large enough to allow many different kinds of visual presentation
• They were seen to be potentially interesting to a general audience
• Between them, the datasets used both numerical and textual data which could be visualised in different ways
• The Observatory had already produced some visualisations using these datasets, which could be compared to the newer, more advanced outputs of Clever Franke.

These bespoke visualisations, along with several others already in public circulation, were used in subsequent focus groups held around Great Britain. At these focus groups, members of the public interacted with the visualisations and shared their thoughts about them. For full details of the methods used, as well as access to resources that aim to improve your confidence with reading visualisations, please see the Seeing Data website.
(http://seeingdata.org/about-seeing-data/what-we-did-on-seeing-data/)

Outputs

Links to the visualisations appear below, and will open a new window. They are research outputs of the Seeing Data project, not final outputs of the Migration Observatory, nor are they currently in use by the Observatory. This is why they appear on the COMPAS website. They should be considered to be work-in-progress versions that are undergoing further refinement based on the emerging results of the AHRC project.

Visualisation : Migration in the Census https://www.compas.ox.ac.uk/migrationinthecensus/

Visualisation : Migration in the News https://www.compas.ox.ac.uk/migrationinthenews/