You are here

New Colloquia Series: Responsible Trust in Science

REASONABLE TRUST IN SCIENCE – DATA SCIENCE

A colloquia series presented by: Science and Technology Studies, and the Public Humanities Centre. These colloquia will be hosted online on Wednesdays, 12:30-2:00. Zoom link information is pending.

Dominic Oldman (ResearchSpace, British Museum): "Building Stronger and Connected Knowledge through Data Driven Cognitive Maps"

Wednesday, January 20, 12:30-2:00

Digital scholarship in the humanities relies heavily on, and is constrained by, the digital architecture of software tools originally designed from a business perspective. The problems humanities scholars face making effective use of these tools lies not with their inability to learn programming skills but with a disconnect between their ‘real world’ contexts of inquiry and the types of structured information environment that computer specialists promote as self-evident, predicated on ontologies that are presumed to be neutral. ResearchSpace is a dynamic knowledge representation system designed by and with humanities and cultural heritage experts with the flexibility to integrate qualitative and quantitative abstractions, and to bring together overlapping data narratives that reflect the diverse vantage points of specific groups and individuals.

Luke Bergmann (Geography, UBC): "Geographical Data Science: Making Spaces for Interpretation”

Wednesday, February 3, 12:30-2:00 

In a world awash, however unevenly, in large spatial datasets being analyzed and theorized at vast scales, what recognition is possible for the situated or interpretative natures of knowledge? How bound up in peoples and places should we expect our understandings of phenomena to be? What roles can the methods of humanists play in the knowledge practices of (geographical) data sciences? Here I examine various efforts to not only resituate associated knowledge claims more modestly, but to rework the associated geographical methods and technologies to facilitate such knowledge projects. In particular, I look at emerging data practices, statistical practices, and cartographic practices that, however differently, grapple with peoples and places as constitutive to computational knowing.

Of related interest:

Sabina Leonelli (Philosophy and History of Science, Exeter University): “Data Science in Times of Pan(dem)ic” 

Wednesday, February 24, 12:30-2:00

Given the reward system focus on the quantity and short-term impact of scientific results, researchers particularly in fields such as biology, biomedicine, epidemiology & data science are primed to look for low-hanging fruits specific to their existing skills and expertise, without necessarily:

  1. devoting attention towards developing datasets and models for longer-term re-use by multiple stakeholders.
  2. considering diverse types of data sources and how they may relate to each other
  3. reflecting on the broader impact of results
  4. ensuring engagement by relevant stakeholders

In this talk, I argue that these trends may have continued and even magnified during the pandemic, with serious consequences for the reliability and usefulness of the research. I discuss some examples from applications of data science to the analysis of contagion rates and sources, and ways in which data use can be re-imagined to offset the shortcomings and instrumentalization confronted by some such projects. I then argue that one way to mitigate this risk is for researchers to  recognise that biomedical and epidemiological expertise needs  to be complemented by other research perspectives (including from social science and humanities), comparisons with other locations/studies, as well as  non-scientific – yet relevant – expertise such as derived from community engagement. I conclude that emergency science can be fast, but should never be rushed; and the need to allow for interdisciplinary, multi-stakeholder consultations - supported by long-term investment in related venues and infrastructures - is heightened when results carry significant public health implications.