Lecture 2

2019

Jan 22, 2019: First lecture

2018:

Jan 22, 2018

reading

• Wallach, Hanna. Big data, machine learning, and the social sciences: Fairness, accountability, and transparency. Medium. Retrieved December 20, 2014, from https://medium.com/@hannawallach/big-data-machine-learning-and-the-social-sciences-927a8e20460d

• boyd, danah, and Kate Crawford. 2012. "Critical questions for big data: Provocations for a cultural, technological, and scholarly phenomenon." Information, Communication & Society 15.5: 662-679. http://www.tandfonline.com/doi/abs/10.1080/1369118X.2012.678878

2017 discussion

• Wallach

  • "uncomfortable": why?
  • types of analyses:
    • when might we describe vs predict vs explain?
    • who might be more interested in one than the other? (e.g., scientists, social scientists, engineers, social media or advertising companies....)
    • how might these be used?
  • "technology": a technology has several implications, e.g.,
    • design choices
    • additional capabilities
    • original intentended purpose of this capability
    • diversity of unanticipated uses of this technology

• boyd & crawford

  • why do we strive for "objectivity"? (the answer should not contain the word "truth")
  • think though the sources of subjectivity in the chronology of you relationship with a dataset, e.g.,
    • how the data were generated
    • your mental model of this
    • choices of what to keep and discard (examples, features)
    • choice of model
    • assessment of model
    • development of model into a technology
    • anticipated uses of this technology
    • unanticipated uses of this technology
  • think of a case where data are not "random" or "representative?"