15 Matching Annotations
  1. Mar 2025
  2. osf.io osf.io
    Critical Quantitative Literacy: An Educational Foundation for Critical Quantitative Research
    6
    1. nicbroschro 22 Mar 2025
      CRITICAL QUANTITATIVE LITERACY8traced back to its roots with scholars of color in critical legal studiessuch as Derrick Bell, KimberléCrenshaw, and Mari Matsuda(e.g.,seeMatsuda et al., 1993).Before them, DuBois (1899)was applying quantitative researchmethods toquestions around racial equity. As a framework, CRTtells us that race is a social construct,and that racism is embedded in legal policies and other social systems.A corollary that race is not readily quantifiable, and quantitative researchinvolving race ought to be critical toward its treatment ofraceand interpretation of its conclusions.

      when conclusions are made based on race, CRT says that race isn't the cause of the difference, but other factors such as access, socioecomnic, etc. come into play.

    3. nicbroschro 22 Mar 2025
      For decades, student grades

      Is state testing another form of this. I know they are culturally non-responsive, but I hadn't put them in this same vain until now.

    4. nicbroschro 22 Mar 2025
      was to utilize the “objectivity” and access to “truth” provided by the mathematical sciences to“prove” the racial and cultural superiority of European

      The basis of the practice was to use data to make decisions and justify choices one couldn't justify without the data. Has always had nefarious intent. Has always taken the human out of the equation.

    Visit annotations in context

    Annotators

    URL

    osf.io/preprints/psyarxiv/346zp_v1
  3. Jan 2025
  4. docdrop.org docdrop.org
    (FFLCH) LIVRO Weapons of Math Destruction - Cathy ONeal.pdf
    5
    2. nicbroschro 30 Jan 2025
      Washington, D.C., schools, to return to that example, evaluates teacherslargely on the basis of students’ test scores, while ignoring how much theteachers engage the students, work on specific skills, deal with classroommanagement, or help students with personal and family problems. It’soverly simple, sacrificing accuracy and insight for efficiency

      I like this example. Test scores compare one thing, but if socio-economics, number of parents in the household, number of siblings, aren't considered, the playing field isn't even.

    4. nicbroschro 30 Jan 2025
      formalize my model,

      This is something I have trouble with at work. I hold a lot of models in my head, but allowing others into the system is really difficult for me. I have never seen it spelled out like this. Good awareness

    5. nicbroschro 28 Jan 2025
      So they substitute stand-in data, or proxies.They draw statistical correlations between a person’s zip code or languagepatterns and her potential to pay back a loan or handle a job. Thesecorrelations are discriminatory, and some of them are illegal.

      The transparency, quality, and level of bias on the data going in affects the quality and bias of the data coming up. Loose coorelations will yield skewed data and predictions.

    Visit annotations in context

    Annotators

    URL

    docdrop.org/download_annotation_doc/-FFLCH-LIVRO-Weapons-of-Math-Destruction---Cathy-ONeal-ti0vb.pdf
  5. docdrop.org docdrop.org
    Ledford-H.-2019-.-Millions-of-black-people-affected-by-racia-dwc2c.pdf
    4
    2. nicbroschro 27 Jan 2025
      This type of study is rare, because research-ers often cannot gain access to proprietary algorithms and the reams of sensitive health data needed to fully test them

      This begs the question about the need for open access and transparency for corporations.

    3. nicbroschro 27 Jan 2025
      The cost model is just one of many data elements intended to be used to select patients for clinical engagement programs

      When the systems that produce these algorithms don't consider bias in their variables, then bias gets perpetuated. For instance, if they believe that cost of care is equal across demographics, they don't believe their results are biases, but their assumptions are off. This seems to be how companies spin their lack of bias.

    4. nicbroschro 27 Jan 2025
      When Obermeyer and his colleagues ran rou-tine statistical checks on data they received from a large hospital, they were surprised to find that people who self-identified as black were generally assigned lower risk scores than equally sick white people

      I am sure many doctors are unaware that they do this, but it is caused by unconscious bias. Exposing it with data unearths it, much like Darden's boss in last article. This shows the need for data like this and that individual cases don't always dispel bias.

    Visit annotations in context

    Annotators

    URL

    docdrop.org/download_annotation_doc/Ledford-H.-2019-.-Millions-of-black-people-affected-by-racia-dwc2c.pdf