We will use a sequential, mixed methods approach in which each stage feeds into subsequent stages: (1) a review of literature and evidence; (2) the ‘data journeys’ method; (3) a survey of knowledge and attitudes; and (4) focus groups and interviews.
Stage 1: Literature Review (1st Oct – 31 Mar 2020).
We have reviewed the available literature than can help us understand how to best answer our research questions: RQ1 What do different people know and feel about specificdata-related practices in different domains of everyday life? and RQ2 What do fair data practices look like, from non-experts’ perspectives?. The review involved:
- a systematic search of online citation databases, focussing on Web of Science. We searched using multiple keywords relating to how people feel about data practices and what happens to their personal data.
- a manual search that began with grey and academic literature with which we were already familiar, and then snowballing out (eg searching bibliographies, relevant websites, observing Twitter discussions and building on word-of-mouth recommendations).
We have written a long report and shortened summary (available soon). They show the outcomes of Stage 1 research which reviewed academic and grey literature published between 2015 and the end of 2019. The document synthesises existing evidence and evaluates whether patterns or generalizable findings emerge from extant research. It assesses the claims that are made on the basis of existing research and reviews methods, analyses and reliability of findings. It identifies limitations and gaps in the existing research that future research can address. This review will inform the original empirical research to be undertaken in the further stages below.
Stage 2: Data Journeys (1st Jan – 31 May 2020).
We have adapted the innovative data journeys approach Bates developed, which produces understanding of specific, concrete data practices by mapping data flows across data infrastructures. The method involved collecting information about data practices and flows in a variety of ways, for example by analysing organisational documents (eg terms and conditions, privacy policies) and interviewing key informants. It did not require access to the sensitive personal data held by our case organisations. This method was applied to our example cases to map some of the ways in which data are captured, processed, shared and used. Because data practices are dynamic and changing, and those undertaken at the time of using the data journeys method may be discontinued in the future, we will use this method to make visible selected data practices at particular moments in time. This approach will result in information about precisely what happens to data in our three cases, enabling us to identify specific data-related practices in different domains of everyday life and to explore what different people know and feel about them (RQ1). We have presented the outcomes of our data journeys work in visual and accessible forms to participants for the next two stages of the research.
Stage 3: Survey (1st May – 30 Nov 2020)
We have carried out a survey (n ≈ 2000), that built on the work done in stages 1 and 2. As noted above, we oversampled members of disadvantaged populations whose knowledge, experiences and perceptions of data practices are important to our focus on what different people know and feel about data practices. These related to demographic variables including household income and educational qualifications, race and ethnicity, age, disability, gender and sexuality.
Survey content covered two main areas. First, we gauged knowledge about specific data practices at given moments in time in the three test cases (and possibly in other examples), by soliciting true/false/don’t know responses to factual statements about case organisations’ data practices identified in Stage 1 and drawn from elsewhere. This addressed the first half of our research question: What do different people know about data-related practices in different domains of everyday life? The survey also gauged attitudes to factual statements, by soliciting responses on a scale (eg from ‘this concerns me’ through ‘I have no opinion’ to ‘this is a good thing’) to the same factual statements and to alternative data practices. This approach has proven effective in previous research (eg Kennedy et al 2015, Turow et al 2015) and enabled us to address the second half of our research question: What do different people feel about specific data-related practices in different domains of everyday life? The survey also provided insights in relation to our question What do fair data practices look like, from non-experts’ perspectives?, although this question will be addressed more fully in Stage 4. Open text fields inviting respondents to explain their responses to attitudinal questions were included in the survey.
Stage 4: Focus groups and interviews (1st October 2020 – 31st May 2021)
We are currently carrying out nationwide focus groups and interviews to explore what people know and feel about data practices in-depth, the types of data practices that they consider to be ‘fair’ and what fairness means to them, building on the survey findings. Due to COVID-19 restrictions, focus groups and interviews are taking place online or over the phone (see our resource for online qualitative research on inequalities here) . Focus groups are taking place across the UK and with groups of people who are already known to each other (to facilitate group meaning-making). Through discussing individual experiences and perceptions of data practices in focus groups participants can gain new forms of understanding which may influence their own future data-related practices (Oman 2017). We are also carrying out nationwide interviews in order to gather both collectively and individually produced qualitative data and to allow us flexibility if assembling groups is difficult or if matters of privacy are raised by potential participants – these interviews may also produce personal knowledge for participants.
A cyclical approach is being used to gather qualitative data, which has proven successful on previous research into non-expert perspectives on data-related issues (Kennedy et al 2015):
- We identify what participants already know and feel about data practices in the example cases.
- We explore what they feel about data practices with which they are not already familiar and which they come to know about through their involvement in the research.
- We present participants with and gauge attitudes to purportedly ‘fairer’ future data practices.
Focus groups and interviews focus on the second part of the question: What do different people feel about specific data-related practices in different domains of everyday life? and our second question: What do fair data practices look like, from non-experts’ perspectives?