Living With Data Literature Review
Home > Resources & publications >
Living With Data: knowledge, experiences and perceptions of data practices carried out an extensive review of existing empirical research into public understanding and perceptions of data practices which was published on 21 May 2020. The full report, a short accessible summary, and accompanying list of references are all available from this page in the Current Research section of this site.
Interest in how the public perceives ‘data practices’ has grown recently. The emerging literature on public understanding and perceptions of datafication needed reviewing, to help us identify next steps for research in the field. We needed to synthesise existing evidence, evaluate whether patterns emerge, assess the claims that are made on the basis of existing research, and identify limitations and gaps that future research can address.
Living With Data therefore undertook such a review. We found:
1. People have some knowledge and understanding of data practices. Findings from both quantitative and qualitative studies suggest that people’s knowledge about what happens to their data is mixed. Knowledge and understanding of data practices are varied. Some people understand some data practices, and they interpret their understanding in different ways which, in turn, leads to different levels of concern.
2. There is extensive evidence across all literature that people are concerned about data practices. This is an important finding that emerges from a lot of diverse research. Policy-makers and data practitioners need to be willing to address these concerns.
3. But this is not the whole picture: people are not only concerned. They find ways to negotiate, embed or resist data practices in their everyday lives. People often hold contradictory views about data practices, recognising their benefits and feeling concerned about potential harms at the same time. In some contexts, people feel they have some agency around their data, especially personal data that they can easily access, such as health self- tracking data.
4. Emotions play an important role in understandings and perceptions of data practices. Qualitative studies recognise the significant role that feelings play in perceptions of data practices. How emotions matter varies across demographic groups. Emotions inform and are informed by reason and rational thinking, so they need to be understood as an important element in public understanding and perceptions of data practices.
5. People trust some sectors with their data more than others. The relationship between trust in institutions in general and trust in institutions’ data practices is complicated, and findings are contradictory. For example, people trust the police with their data but they do not trust automated, data-driven decision-making in criminal justice practices. Qualitative research suggests that trust and distrust in data practices are not experienced separately. Trust and distrust are context-dependent, and sometimes trust and distrust co- exist. Sometimes, distrust is appropriate, because trust needs to be earned. Research suggests that people believe that better communication and the existence of safeguards, accountability and transparency would make organisations more trustworthy.
6. Some responses to data practices are seen as apathy or acceptance. But responses need tobe understood in a context in which people feel unable to control the flows of their personal data, even if they want to. Some researchers see this as digital resignation, not apathy or acceptance. Using data-driven services does not mean that people accept data practices – we also saw that people are concerned about them and that people hold contradictory views about them. In addition, sometimes people resist data practices, and there are various ways in which they do this.
7. Most research finds dissatisfaction with the current ways in which data is used and managed, and a desire for this to change. A number of characteristics of changed, fairer data practices have been identified. These include:
- Honesty, transparency and genuine dialogue with the public;
- Regulation, enforcing compliance, the existence of safeguards and accountability, and the right to redress;
- Personal control.
As with people’s concerns and degrees of trust, contexts of data use influence people’s thinking about whether they are fair or not.
8. Because people are concerned about existing data practices, because trust in them is limited, and because people have views about how data practices could be fairer, clearly, change is needed. Views about what needs to change are influenced by research discipline and researcher perspective. Systems-focused literature recommends changes in system design, foregrounding actions that could be taken by technology providers, such as clear communication, privacy by design, and attending to complex privacy and contextual dynamics. Critical academic literature and some policy and practice-oriented grey literature identify governments as the most important enablers of change.
9. Differences matter when it comes to public perceptions of data practices. Not all data
practices are the same, and people experience them from different social positions. Research is beginning to pay attention to differences, but more understanding of them is needed. Differences matter in relation to types of data or contexts of data use and how concern about data practices differs in significance from other concerns that people have. Most importantly, social inequalities play a major role in shaping people’s experiences of data practices, and therefore their understanding and perceptions of them.
10. How research is conducted makes a difference to what it finds. Research methods, the questions asked, how findings are interpreted and presented, the disciplinary background and the political orientation of researchers all play a role in shaping findings that emerge and claims that are made in the research we reviewed. Decision-making based on the evidence we have reviewed should be alert to this fact.
Three overarching conclusions emerged:
I. Data matters are human matters. This means that data-related governance
and decision-making needs to be human-centric. It needs to start with the experiences and perceptions of the people who are affected by data practices.
II. Context matters.Who gathers data, what and whose data is gathered, for what purpose and with what effects, influences people’s attitudes.
III. Inequalities matter. Social inequalities influence knowledge and understanding, concerns, degree of trust and feelings about data practices.