24 Matching Annotations
  1. Sep 2016
    1. generally good

      Having read through these sections now I am struck by the absence of any discussion of what geographers call "volunteered geographic information" and the potential that "big data" has for mapping social phenomena because data points often are frequently geo-referenced. Also, these data tend to offer broader geographic coverage than many traditional data sources.

    2. inaccessible


    1. is sensitive.

      Some of it is sensitive, some has the potential to become sensitive (e.g. when a person who's been sharing location data starts being stalked, previously harmless-seeming data becomes sensitive).

    1. Although bigness is generally a good property when used correctly, I’ve noticed that bigness commonly leads to a conceptual error. For some reason, bigness seems to lead researchers to ignore how their data was generated. While bigness does reduce the need to worry about random error, it actually increases the need to worry about systematic errors, the kinds of errors that I’ll describe in more below that arise from biases in how data are created and collected. In a small dataset, both random error and systematic error can be important, but in a large dataset random error is can be averaged away and systematic error dominates. Researchers who don’t think about systematic error will end up using their large datasets to get a precise estimate of the wrong thing; they will be precisely inaccurate (McFarland and McFarland 2015).

      this is helpful

    2. over time

      Strengths of having time series is not really discussed below, but are evident in the context of event detection and things like that.

    1. administrative

      perhaps "transactional" would be a better word than administrative

    2. a study that I’ll tell later in this chapter

      awkward phrasing

    3. by governments

      why by governments? private actors, too

    1. power

      Probably premature to raise this, but in my view it's not just about increasing power but also about a shifting balance of power between researchers, research informants, and the commercial entities that own/control data (which didn't use to be in the picture)

    2. cheaper

      more cheaply

    3. start interacting

      starting to interact

    1. will

      that seems optimistic -- there are lots of new practical considerations in connection with digital age research (such as data management), I wouldn't expect researchers to automatically be freed up to think about ethics. Their ability to be ethical has a lot to do with their work conditions, which both within and outside of academia aren't always favorable.

    1. The best place to start is research design.

      I think one step before -- how to ask questions -- is also quite an important one for both communities to work on together, but I see your point that connections can be forged more easily around practical concerns than around abstractions and theories.

    1. show no sign of slowing down

      Not sure about that -- there's been talk of an end of Moore's law and that the right-hand curve will start looking more like a sigmoid curve (on this, I'm actually thinking of some musings by Peter Frase: http://www.peterfrase.com/2010/12/social-science-fiction/)

    2. combine social science with data science

      Here you make it appear like this combination is mostly about combining different kinds of methods, but in the preface you alluded to "wisdom" of social thought in making sense of social data. That is absent from the discussion so far.

    3. more of the same becomes something different

      "the dialectic of quantity and quality"

    4. information

      why go with "information" rather than data? Information implies that this data deluge somehow has the capacity to inform, which may or may not be the case

    5. digital phone

      awkward, better to say smart

    1. It is for social scientists that want to do more data science, and it is for data scientists that want to do more social science.

      Will most sociologists picking up a book like this know what "data science" is? I have difficulty estimating how mainstream this term has become.

    2. distinction


    1. I went and looked up the article because I immediately had a million questions that aren't addressed in your summary of Blumenstock et al.'s research. At first I was going to suggest providing some additional detail (for instance about why mobile phone records can be assumed to provide good coverage in a country like Rwanda), but then I realized this is probably exactly the right kind of feeling to leave readers with at this point: eager to find out more about how it all works. So I think this section is quite effective.