Creative data literacy - Bridging the gap between the data-haves and data-have-nots

growing gap with those who can or can’t effectively work with data
“it is state and corporate actors who possess the resources to collect, store and analyse data”
individuals are more likely to be subjects of data than to be able to use it for their gain
“Literacy, following the model of popular education proposed by Paulo Freire, requires not only the acquisition of technical skills but also the emancipation achieved through the literacy process.” (Freire 1968; Tygel & Kirsch 2015).
proposes the term ‘creative data literacy’ to refer to non-technical data fields
creative data literacy for empowerment
“working towards creative data literacy is not only the work of educators but also of data creators, data publishers, tool developers, tool and visualization designers, tutorial authors, government, community organizers and artists”
“The future is already here. It’s just not very evenly distributed.”– William Gibson

‘profound inequality’ between thise who benefit from Big Data (see Andrejevic 2014; boyd & Crawford 2012; Tufekci 2014)
‘Data has become a currency of power’
publicly relevant decisions are made increasingly by automation over huge amounts of data. (see Pasquale 2015)
data ownership is centralised, owned by corporations and govs
data management knowledge held only by specialists
“People are far more likely to be discriminated against with data or surveilled with data than they are to use data for their own civic ends (O’Neil 2016).”
Critical Data Studies often focuses on algorithms and privacy over who has access to computing power and technical knowledge to make sense of and deploy data
Big Data divide
equity and inclusion issues in data science as much as any other technical field
“Data literacy includes the ability to read, work with, analyze, and argue with data as part of a broader process of inquiry into the world (D’Ignazio & Bhargava 2016; Letouzé et al. 2015).”
various workshops and social initiatives on data for self-advocation but little cohesive effort to get novices to ‘speak data’ (Bhargava 2014) or develop a “data mindset” (Miller 2014).
“open data movement” (Gray 2012).
this paper offers five tactics for creative data literacy for empowerment.
not only the domain of the educator, but various groups of professionals - data creators, data publishers, tool developers, tool and visualization designers, tutorial authors, government, community organizers, artists.
need a research agenda to “cultivate data literacy for empowerment across numerous sectors, including and especially “the accountability industries”: Law, Government, Journalism, Education and the Arts.” accountability industries

community-centered data

“careful sourcing and selection of data that are relevant to the community that is learning to work with data.”
not generic sample data but data directly relevant to issues faced by learners
“Working with community-centered sample data opens up possibilities for connecting context and lived experience to the data. It also makes it easier for learners to apply their learning to their everyday lives or work contexts more quickly and directly.”
“They had collected the data, they had deep, ongoing everyday relationships with the people and the place and, most importantly, they had a stake in the outcome of the data analysis.”
some good examples here of relevant contextual data for learners
fun and varied datasets sometimes work, but “can sometimes fail to connect to learner’s more pressing professional contexts.”

data biographies

data in the wild
open data avoids extra admin eg having to request specific public records
but “data arrives at our doorstep completely decontextualized, without explanation of why it was collected, who collected it, in what way, and what its known limitations are”
good open data has data dictionary/playbook/metadata to learn about the data’s provenance, but many publishers don’t have the resources to do this
“new learners tend to see information organized systematically in a spreadsheet as ‘true’ and complete”
“ask learners to write ‘data biographies’. These are stories of how the data set came to be in the world.”
“estimating whether patterns in the data are an artifact of the collection process or a signal in themselves” signal vs noise
“Creating a data biography might be as simple as invit ing the owner of a data set to present about the process of collection, or encouraging learners to interview one of the creators or maintainers of a data set. Occasionally for this assignment, learners have found that their data biography has turned into the whole story.”

make data messy

“creating and categorizing data in the face of uncertainty and complexity”
“Cultivating skepticism of “raw” data, therefore, should be seen as one of the primary goals of any data literacy program that seeks to empower the learners.” - data not inherently truth people assume truth science is all theories people assume data is true
“Taking the learners through the process of data collection, categorization and standard-creation helps them understand how inquiry goals, interests and politics contribute to the creation of data sets. Furthermore, it helps learners engage the critical thinking skills they will need to ask questions of other data sets i”
“to cultivate skepticism of raw data, especially data collected about the physical world by technological instruments.”
“Making data messy can also be employed by institutions working with diverse stakeholders and constituents as a way to engage in collaborative analysis and meaning-making from data. […] While a data analyst could have been hired to produce an analysis in the shadows of city hall, the project leadership instead treated the analysis as an opportunity for further community engagement.”
making data messy does not mean it has to stay messy.
“The challenges that result are epistemological (i.e., how can we gather data to make a claim about the world?) and editorial (i.e., what do we need to include/exclude?).”

Build learner-centred tools

many tools with varying purpose and little curation for purpose, especially whilst learning
many prioritise flashy visuals over “scaffolding a learning process that helps the learner through each stage of the data processing pipeline.”
“A learner-centered data literacy tool is:
1. Focused: Strives to do one thing well. Provides a low barrier to entry for the data literacy learner. Each tool in the DataBasic suite is focused on taking in one type of input and producing a single web page report or visualization as output.
2. Guided: Introduced with strong activities and sample data to get the learner started so they do not have to imagine use cases. The input screen of each DataBasic tool starts with sample data so the learner can quickly run it to see what kind of output it generates.
3. Inviting: Appeals to the learner—either because of direct relevance to the learner themselves or through the use of play, humor and visual design. DataBasic’s visual design uses bright colors and simple layouts. T he sample data draw from pop culture, something the user may be familiar with.
4. Expandable: Helps the learner take the next step (possibly to another, more advanced tool). Each tool in the DataBasic suite recommends two other tools that learners can use once they are ready to take the next step”
“learner-centered tools must be expandable and help the learner “graduate” to more open-ended tools.”

Favor creative, community-centered outputs over Tuftean purity

outside of data visualisation, “the range of outputs produced should rightly expand to accommodate people’s increasingly diverse perspectives, goals and situations”
“precise comparison may be goals if one’s audience consists of scientists or designers, such graphical language might not be appropriate” for other community groups
“Journalists, artists, citizen scientists and makers are experimenting with ways of visualizing, physicalizing or even ‘visceralizing’ data in order to more effectively communicate their data-driven ideas.”
mapping as data visualisation