Diverging Data: Exploring the Epistemologies of Data Collection and Use among Those Working on and in Conflict

ABSTRACT

The emergence of ‘big data’ and calls for ‘evidence-based decision-making’ to increase the effectiveness of peacekeeping, humanitarian, and development programming have resulted in an often-unadulterated embrace of the promise of data as a solution to a diverse set of problems facing these sectors. At the same time, however, the increasing and widespread collection of data – by researchers, policymakers, and operational actors – directly contrasts with the sense that data are often collected yet not used. Taking a step back, this article examines the epistemologies of data collection and use from the perspectives of the scholar/researcher, and the practitioner/operational actor to illustrate how these perspectives elicit a series of data divergences. I argue that the collection and use of data replicate and are poised to extend the theory-practice divide that exists between researchers who study violence – those working ‘on’ conflict – and the peacekeepers, peacebuilders, and aid workers who work ‘in’ the midst of it. Differing conceptions of the purpose and use, sources and characteristics, and time frames of data reflect role-based positionalities that shape practices of collection, affect their interoperability, and limit their possible use, even as combining these types of data could also address their inherent limitations.

Acknowledgements

I thank Roger Mac Ginty and reviewers for their useful feedback on an earlier version of this article, Will H. Moore and colleagues at HCRI (Manchester) and DPCR (Uppsala University) for influencing my thinking, and the US-EU Fulbright program, since I wrote this article while a Fulbright-Schuman Research Scholar.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Larissa Fast (PhD) is Senior Research Fellow at the Humanitarian Policy Group (London) and Research Fellow with the IIHA, Fordham University. Fast's research has focused on humanitarianism, peacebuilding, risk and security from the perspective of one who has worked on and in conflict.

Notes

1 Ericsson, Mobile Subscriptions and Penetration. This exponential growth and the expanding access to the Internet nevertheless masks dramatic inequalities, based on region, gender, and income (World Bank, World Development Report).

2 ITU, ICT Facts and Figures 2016.

3 UNICEF, “Mobile Apps Are a Lifeline for Young Syrian Refugees.”

4 The PetaJakarta website (https://petajakarta.org/banjir/en/, accessed 12 December 2016) uses hashtags and tweets to create real-time maps of flooding. PRI, “Need the Latest News on Flooding?”

5 One example is Mapswipe. The tagline of MapSwipe, associated with the Missing Maps project (http://mapswipe.org and www.missingmaps.org, accessed 12 December 2016) is ‘Humanitarian organizations can't help people if they can't find them. Mapswipe is a mobile app that lets you search satellite imagery to help put the world's most vulnerable people on the map.’

6 I use the term peace support operations as an inclusive term, referring to operations conducted in complex security environments that encompass peacekeeping, peace enforcement, as well as peacemaking, conflict prevention, peacebuilding, humanitarian, and development interventions. See NATO, Peace Operations Support. For the contrasting perspective, which argues against the inclusion of humanitarian action under the broader rubric of peace support operations or even development or peacebuilding, see DeTorrenté, “Humanitarianism Sacrificed.”

7 DeWaal et al., “The Epidemiology of Lethal Violence”; Bradt, “Evidence-based Decision-making”; and Ramalingam, “Case Study.”

8 Salehyan, “Best Practices in the Collection of Conflict Data”; Gleditsch, Metternich, and Ruggeri, “Data and Progress in Peace and Conflict Research.”

9 I use these as archetypes of sorts, since some practitioners read academic journals and some academics actively engage practitioners and policymakers. Still others function as bridges between these two worlds. Nevertheless, they tend to function as exceptions who prove the rule.

10 Goodhand's distinction refers to the goal of peacebuilding, where working ‘on’ conflict is to work for positive impacts on peace dynamics, whereas working ‘in’ conflict aims to limit the negative impacts of programming on conflict dynamics. Goodhand, “Working ‘In’ and ‘On’ War.”

11 I use epistemologies to refer to the ‘creation and dissemination of knowledge’, and especially its sources and limits. Stanford Encyclopedia of Philosophy, “Epistemology.”

12 I do not conflate policymakers and practitioners, since they collect and use data for different purposes (e.g. action vs. advocacy) but do not delineate these differences in the article. These differences are apparent from the tensions between these groups that sometimes afflict operational organizations that both implement their own programmes and engage in policy advocacy work. See HPG, “Humanitarian Advocacy in Darfur.”

13 Mac Ginty, “Routine Peace.”

14 OCHA, “Humanitarianism in the Network Age, 31”; Meier, Digital Humanitarians; HHI, “Disaster Relief.”

15 OCHA, “World Humanitarian Data and Trends 2016,” 3. An interactive version is available at www.unocha.org/datatrends2016/, accessed 15 December 2016.

16 UN, “Agenda for Humanity,” www.agendaforhumanity.org, accessed 27 January 2017.

17 OCHA, “Humanitarian Data Exchange,” https://data.humdata.org/, accessed 15 December 2016.

18 Sachs, “Data for Development.”

19 Flowminder, “Case Study.” Such initiatives and other uses of digital technology compose what is known as digital humanitarianism (Meier, Digital Humanitarians). For another example of technology-enabled, real-time data collection and the challenges of digital – vs. paper-based data, see Evidence Action, “Going Paperless.”

20 See Derbyshire and Donovan, “Adaptive Programming in Practice”; Mercy Corps, “Managing Complexity.” Adaptive management (AM) is also known as adaptive programming.

21 In full disclosure, my former close colleagues at USAID have been involved in promoting AM within USAID and across the donor community. For more on AM, see Green, “Adaptive Management is Here to Stay”; Green, “Doing Development Differently.”

22 See, for example, the Doing Development Differently manifesto (http://doingdevelopmentdifferently.com, accessed 16 December 2016) and the calls for reform that accompanied the May 2016 World Humanitarian Summit (e.g. Bennett, “Time to Let Go.”)

23 Data are both numeric (usually referred to as quantitative) and text based (narratives, stories, interviews). Both quantitative and qualitative researchers collect and analyse data, although quantitative scholars more often publicly release their data sets.

24 E.g. Kalyvas, The Logic of Violence in Civil War; Downes, “Desperate Times, Desperate Measures”; Herreros and Criado, “Pre-emptive or Arbitrary”; Stanton, “Regulating Militias.”

25 Lacina and Gleditsch, “Monitoring Trends in Global Combat.”

26 Pettersson and Wallensteen, “Armed Conflicts.”

27 Eck and Hultman, “One-sided Violence against Civilians in War.”

28 Cohen and Green, “Dueling Incentives.”

29 Findley et al., “The Localized Geography of Foreign Aid.”

30 See www.correlatesofwar.org, accessed 15 December 2016.

31 See http://ucdp.uu.se, accessed 15 December 2016.

32 Gleditsch, Metternich, and Ruggeri, “Data and Progress in Peace and Conflict Research.”

33 See www.acleddata.com, accessed 15 December 2016, and www.systemicpeace.org, accessed 15 December 2016 for additional data sets. Although many more data sets exist, I only list several here.

34 See http://fsi.fundforpeace.org/rankings-2015, accessed 15 December 2016.

35 For example, Paige Fortna's data set on Peacekeeping in Civil Wars, www.columbia.edu/~vpf4/research.htm, accessed 15 December 2016 and the in-progress Peacemakers at Risk (PAR), a sub-data set of the UCDP armed conflicts data, www.pcr.uu.se/research/ucdp/program_overview/current_projects/peacemakers_at_risk/, accessed 27 January 2017.

36 See https://peaceaccords.nd.edu, accessed 15 December 2016.

37 Sandvik et al., “Humanitarian Technology,” 225.

38 Read, Mac Ginty, and Taithe, “Data Hubris?” 7.

39 Currion, “Eyes Wide Shut”; Currion, “Game of Drones.”

40 Jacobsen, The Politics of Humanitarian Technology, 41; also Jacobsen, “Experimentation in Humanitarian Locations.”

41 White, “Images of Refugee Camps”; Malkki, “Speechless Emissaries”; Fehrenbach and Rodogno, Humanitarian Photography.

42 Givoni, “Between Micro Mappers and Missing Maps,” 1038.

43 Duffield, “Disaster-resilience in the Network Age”; Fast, “Unpacking the Principle of Humanity”; and Tapscott and Desai, “Tomayto-Tomahto” for a similar take regarding qualitative research in fragile environments.

44 Sandvik et al., “Humanitarian Technology,” 222.

45 Read, Mac Ginty, and Taithe, “Data Hubris?” 1325.

46 Read, “Tensions in UN Information Management.”

47 Van de Walle and Comes, “On the Nature of Information Management in Complex and Natural Disasters,” 407.

48 Currion, “Slave to the Algorithm”; Owen, “The Violence of Algorithms”; Sandvik et al., “Humanitarian Technology.”

49 O’Neil, Weapons of Math Destruction; Le, “Weaponized Data.”

50 Atal, “Patrick Ball on the Perils of Misusing Human Rights Data”; Price and Gohdes, “Searching for Trends.”

51 Salehyan, “Best Practices in the Collection of Conflict Data”; Davenport, Media Bias; Davenport and Ball, “Views to a Kill.”

52 Spagat and Johnson, “Colombia Did Not Vote No” analyses how human error affected counting, and therefore the results, of the Colombian referendum.

53 People in Guinea talked about ‘Ebola business’, referring to the tangible signs of wealth that accompanied the outbreak, such as the 4 × 4 vehicles and new buildings. In interviews, I conducted in Guinea in February 2016 (see below), interviewees reported that back in the spring of 2014 (after the laboratory tests confirming Ebola but well before the outbreak peak), community members calling into radio shows about Ebola asked if white people brought Ebola or if it was a way to funnel money into poorly regarded government projects.

54 Samb, “Anger, Mistrust in Guinea Villages Hinders Battle to Beat Ebola.”

55 Rao, “When Waiting for Data Is a Death Sentence.”

56 Sandvik et al., “Humanitarian Technology”; Jacobsen, The Politics of Humanitarian Technology; Raymond and al Achkar, “Building Data Responsibility into Humanitarian Action”; Raymond and al Achkar, “Data Preparedness”; OCHA, “Humanitarianism in the Network Age”; OCHA, “Humanitarianism in the Age of Cyber-warfare.”

57 Kaspersen and Lindsay-Curtet, “The Digital Transformation of the Humanitarian Sector.”

58 Human Security Report, “The Causes of Peace and the Shrinking Costs of War.”

59 Hultman, Kathman, and Shannon, “UN Peacekeeping and Civilian Protection in Civil War.”

60 Wille and Fast, “Humanitarian Staff Security in Armed Conflict”; Sheik et al., “Deaths Among Humanitarian Workers”; Stoddard, Harmer, and DiDomenico, “Providing Aid in Insecure Environments.” See also www.safeguardinghealth.org, http://healthcareindanger.org, and www.pcr.uu.se/research/ucdp/program_overview/current_projects/peacemakers_at_risk/, all accessed 27 January 2016.

61 See https://aidworkersecurity.org and www.ngosafety.org, accessed 27 January 2017.

62 See http://aidindanger.org, accessed 27 January 2017; OCHA, “Access Monitoring and Reporting Framework.”

63 See www.acaps.org and www.peacebuildingdata.org. Two other academic blogs (http://politicalviolenceataglance.org and www.washingtonpost.com/news/monkey-cage/) are aimed at policymakers. All accessed 27 January 2017.

64 Wille and Fast, “Operating in Insecurity”; Fast and Wille, “Introducing the SiND.” All policy reports available at http://aidindanger.org/analysis, accessed 27 January 2017.

65 de Waal et al., “The Epidemiology of Lethal Violence”; see also other articles in this special issue.

66 Another example is Stanton and Narang, “The Logic of Attacks on Third Parties in Civil War.”

67 Fast and Waugaman, “Fighting Ebola with Information.”

68 In reality, epidemiologists may be researchers or field based and are sometimes both. Adopting my earlier distinction, research epidemiologists work ‘on’ epidemics, often remotely from labs in other countries, while field epidemiologists work in the midst of epidemics, tracking the disease and investigating its causes and transmission.

69 Fast and Waugaman, “Fighting Ebola with Information.”

70 OCHA, “The Humanitarian Data Center in the Netherlands,” 21.

71 Diehl et al., “Ebola Virus Glycoprotein with Increased Infectivity Dominated the 2013–2016 Ebola Epidemic.”

72 This Indian parable is described in a John Godfrey Saxe poem titled “The Blind Men and the Elephant.”

73 Obviously these lines may blur (e.g. some scholars engage in action-oriented research while some practitioners conduct rigorous research focused on the accumulation of knowledge).

74 Some journals (e.g. the Journal of Peace Research) now require researchers to make their data available as a condition of publication.

75 See Davenport and Moore, “Conflict Consortium Data Standards and Practices for Observational Data”; also Gleditsch and Melitis, “Symposium on Replication in International Studies Research.”

76 Eck, “In Data We Trust?”

77 Lacina and Gleditsch, “Monitoring Trends in Global Combat.”

78 Fast, Aid in Danger.

79 Fast and Waugaman, “Fighting Ebola with Information.”

80 Pettersson and Wallensteen, “Armed Conflicts.”

81 Conrad, Haglund and Moore, “Disaggregating Torture Allegations.”

82 Tilly, “Event Catalogs as Theories.”

83 Davenport, Media Bias.

84 Eck, “In Data We Trust?”; Salehyan, “Best Practices in the Collection of Conflict Data.”

85 Eck and Hultman, “One-sided Violence against Civilians in War.”

86 Fazal, “Dead Wrong?”

87 Human Security Report, “The Causes of Peace and the Shrinking Costs of War.”

88 Anderson, Brown, and Jean, “Time to Listen.”

89 See “The Grand Bargain,” www.agendaforhumanity.org/initiatives/3861, accessed 27 January 2017.

90 Clark, “We’re Over-researched Here!”

91 The Joined-Up Data Project (http://juds.joinedupdata.org, accessed 16 December 2016) describes the problem as ‘loads of data and too little information’ caused by myriad data collection processes, formats, and standards, and the solution as ‘joining up data standards’ to enable the interoperability of data.

92 Blair, Blattman, Hartman, “Predicting Local Violence.”

93 See ICPSR Data Archive, www.icpsr.umich.edu/icpsrweb/ICPSR/, accessed 27 January 2017.

94 Credit to Róisín Read for this idea. In her article “Tensions in Information Management”, Read points to peacekeeping data as both a physical and an imaginative archive, which could be used to shape discourses of the conflict long after the data are collected.

95 See discussion in Davey, “Humanitarian History in a Complex World.”

96 Henry, Higate and Sanghera, “Positionality and Power”; Ryan, Kofman, and Aaron, “Insiders and Outsiders.”

97 See https://centre.humdata.org, accessed 18 December 2016.