b'ManagementYuen Yuen Ang is Associate Professor of Political Science at the University of Michigan, Ann Arbor. She is also a faculty associate at UMs Center for the Study of Complex Systems. In 2018, Ang was named an Andrew Carnegie Fellow by the Carnegie Corporation of New York.4.Combine big and thick data to improve communication.structured or unstructured, and the number of dimensions 5.Improve government agencies knowledge of mixedcaptured by the data.research methods. If big data reflects volume, velocity, and variety for items that can be counted, thick data concerns information about the Analytical Perspectives on Big data significance, meaning, and connections that humans assign to and Thick Data services or technologies, as well as the process by which they Big data can be described as massive data sets sifted byconsume them. Thick data is generated through immersion in powerful analytical tools. 1Big data is defined not only byusers natural settings, rather than in laboratory-type settings.volume (the amount of data collected and analyzed), but also by velocity and variety. Velocity refers to the speedTable 1 compares the attributes and functions of big data and of data being produced, for example; streaming data isthick data. Table 2 demonstrates the key characteristics of the now widely available for real-time, instant analysis. Varietythree case studies examined. refers to the different forms of data collected, which can be Table 1: Big data vs. Thick Data Big Data Thick DataFormat of data Data in numerical formatData in non-numerical, qualitative format VolumeUsually large, even massive number ofUsually small number of observationsobservationsDigital records, digitized archival records,Participant observation, direct participant Methods of data collection streaming data, transmission logs, numerical datainterviews, focus groups, small-n open-ended scrapped from the Internet and social mediasurveys, video recordings, qualitative data from the Internet and social mediaAnalysts Research by social and computational scientistsResearch by anthropologists and ethnographersUsually on-site or direct observation online, and Immersion Requirement Analysts need not be on-site to analyze the data immersed in context (observing, participating, talking to people)Generate solutions to largely known problems,Identify problems that matter most to Role in problem solving such as by mapping behavioral patterns, makingstakeholders, particularly unknown problems and predictions, generating automated decisionspreviously unexpressed needs, and test solutions before scaling upScale: generate insights that are generalizable to aDepth: identify what stakeholders care about in Strengthslarge portion of or an entire population the first place; paint a holistic picture of certain experiencesSource: Yuen Yuen Ang1. The Power of Thick Data, Wall Street Journal, March 21, 2014.WINTER 2019 / 2020 IBM Center for The Business of Government 93'