Big–Thick Blending: A method for mixing analytical insights from big and thick data sources

Abstract

Recent works have suggested an analytical complementarity in mixing big and thick data sources. These works have, however, remained as programmatic suggestions, leaving us with limited methodological inputs on how to archive such complementary integration. This article responds to this limitation by proposing a method for 'blending' big and thick analytical insights. The paper first develops a methodological framework based on the cognitivist linguistics terminology of 'blending'. Two cases are then explored in which blended spaces are crafted from engaging big and thick analytical insights with each other. Through these examples, we learn how blending processes should be conducted as a rapid, iterative and collaborative effort with respect for individual expertise. Further, we demonstrate how the unique, but often overlooked, granularity of big data plays a key role in affording the blending with thick data. We conclude by suggesting four commonly appearing blending strategies that can be applied when relying upon big and thick data sources.

Computational social science using topic modeling: Analyzing patients' values using a large hospital survey

“…These results provide a proof‐of‐concept of the utility of this approach for partially automating thematic analysis by using topic modeling. The approach goes beyond previous computational social science studies that have overlaid thematic analysis upon automatically generated dictionaries (e.g., Fleischmann et al, ) and provides evidence of the ability to combine big data with the thick description of Geertz (1973) (Bornakke & Due, ). The results of this study in terms of values resonate with the findings of field studies of what patients value in hospital care (e.g., Joffe et al, ; Morris, ).…”

Hybrid Methodology: Combining Ethnography, Cognitive Science, and Machine Learning to Inform the Development of Context‐Aware Personal Computing and Assistive Technology

“…Ethnography is no stranger to hybrid approaches -for example, anderson et al (2009) have explored the combination of qualitative research with data mining into "ethno-mining [...] a hybrid, not a 'mixed method'; it is two elements that cannot be separated out [...] [yet] traces of each of the ingredients can still be seen -the same ethos of ethnography (open-ended, coconstructed, holistic field research) integrated with the empirical and analytical capacities of quantitative data mining" (anderson et al 2009, 125). Applied social scientists have also been exploring the blending of "big data" with "thick data" (Bornakke and Due 2018) and outlining approaches like "Contextual Analytics: a project process for uniting data analysts and social scientists under the mandate of building more effective and credible algorithms" (Arora et al 2018, 225). Our work hopes to carry this thinking forward.…”

Innovating Metrics for Smarter, Responsive Cities

“…The primary conclusion of this paper highlights the potential for developing new pathways Encouraging an ethnographic understanding of data, Dourish and Cruz [28] argue for a narrating of and with data "as part of broader processes of interpretation and meaning-making" because it is claimed that "data do not stand alone" but rather, "they do their work in relation to multiple other entities" whether this be aggregation, systems of processing, or "most importantly in relation to people". As if in response, Bornakke and Due [29] propose a method for big-thick blending in support of analytical insights such that the data universe is conceptualized on a horizontal range of thin to thick and a vertical range of extensive to small. As such, big-thin data from surveys can be plotted in the upper left and small thick data from interviews and observations in the lower right [29] in the generation of "unique complementary effects".…”

“…As if in response, Bornakke and Due [29] propose a method for big-thick blending in support of analytical insights such that the data universe is conceptualized on a horizontal range of thin to thick and a vertical range of extensive to small. As such, big-thin data from surveys can be plotted in the upper left and small thick data from interviews and observations in the lower right [29] in the generation of "unique complementary effects".…”