Leiter Reports: A Philosophy Blog

News and views about philosophy, the academic profession, academic freedom, intellectual culture, and other topics. The world’s most popular philosophy blog, since 2003.

  1. Steven Hales's avatar
  2. sahpa's avatar

    Essays as coursework has never been just about engaging the argument itself. Authorship matters because it matters that the argument…

  3. André Hampshire's avatar

    If anything, this exchange illustrates the problem: judgments are being made on stylistic impressions (“this sounds like AI”) rather than…

  4. Ted Bach's avatar

    The existential threat is not to higher-ed as such but a particular (and now common) higher-ed business model: the one…

  5. Steven Hales's avatar
  6. Collin Lucken's avatar
  7. André Hampshire's avatar

    Sagar’s claim that LLMs pose an “existential threat” to universities rests on a set of conflations that do not survive…

More evidence that social distancing is working: Kinsa thermometers recording fewer fevers

Brian Leiter

The NYT ran a story on this today, and the Kinsa map is here.  From the NYT story:

The company, Kinsa Health, which produces internet-connected thermometers, first created a national map of fever levels on March 22 and was able to spot the trend within a day. Since then, data from the health departments of New York State and Washington State have buttressed the finding, making it clear that social distancing is saving lives….

Kinsa’s thermometers upload the user’s temperature readings to a centralized database; the data enable the company to track fevers across the United States….

Kinsa has more than one million thermometers in circulation and has been getting up to 162,000 daily temperature readings since Covid-19 began spreading in the country….

To identify clusters of coronavirus infections, Kinsa recently adapted its software to detect spikes of “atypical fever” that do not correlate with historical flu patterns and are likely attributable to the coronavirus.

As of noon Wednesday, the company’s live map showed fevers holding steady or dropping almost universally across the country, with two prominent exceptions [in New Mexico and Louisiana].

Leave a Reply to Kenny Easwaran Cancel reply

Your email address will not be published. Required fields are marked *

3 responses to “More evidence that social distancing is working: Kinsa thermometers recording fewer fevers”

  1. A in Sahron

    Prof. Leiter, isn't it in the interests of this private company to announce they have a system in place to track the viral spread? I don't know if their data is good or not. Are they willing to open up their software changes to public scrutiny? How do we know they are not modifying the software to match what is already known about the current hot spots? Do they have a public interest or an interest in selling more thermometers? Just some food for thought.

    BL COMMENT: The NYT claims to have received and examined the underlying data. They also have a track record of doing this with flu, where they are able to predict what CDC data confirms a few weeks later.

    Reply
  2. Kenny Easwaran

    I'm worried about how little they have said about the details of their data. It sounds like the data are generated by thermometers owned by households, reporting their data to the company that makes the thermometer. It seems like reported temperatures would go down if more healthy people start measuring their own temperature every day, rather than just measuring their temperature on days they already feel sick (which might have been the normal behavior). It's possible they've got some smart way to correct for this effect, but the extent of what they say in their FAQs is:

    "As of March 2020, we are seeing 2-3x the number of users taking temperatures than we've tracked in previous flu seasons. This does not impact our illness signal, as our modeling already accounts for rapid changes in our user base. We also benchmark our signal against the Centers for Disease Control (CDC) at the end of every flu season when the CDC has finalized their illness reporting. We regularly see an in-season correlation of R2 of >= 0.95.

    Key Takeaway: A significant and rapid change in temperature readings does not affect our illness signal."

    I still worry about other potential confounds, like multiple household members taking their temperature with the same thermometer without properly indicating who is who. It's possible that the households that bought these thermometers are unusually skilled at correctly using devices like this, and unusually diligent about measuring their temperature, but if so, then their measurements are likely to be unrepresentative of the county in other ways (such as being from populations that are disproportionately likely to obey public health orders, and disproportionately likely to have had travel habits that would lead them to be "early adopters" of the infection relative to the rest of their county).

    Reply
  3. srv

    Somewhat OT:

    https://www.bloomberg.com/opinion/articles/2020-03-29/coronavirus-pandemic-puts-moral-philosophy-to-the-test

    One hopes, philosophers will do better than end up playing soccer!

    —–
    KEYWORDS:
    Primary Blog

    Reply

Designed with WordPress