Thursday, May 21, 2020
We are starting to see a more constant and robust stream of data around many topics related to COVID. And, as these data are published or shared, each one needs to be thought about critically - what were the data gathering biases? What do the analyses demonstrate? Are the conclusions logical or marred by fallacy?
And, speaking of data, I got a surprising number of responses talking about individual experiences with robot Vacuum cleaners (Dustin' has cousins, it appears). Amongst the various data points, I learned that 2/2 respondents who name their robots name those robots, Rosie. Next line of investigation - why do we apply genders to robots, and why does the Jetson's robot inform the naming of these respondents? A fascinating chain of thought reflecting a mélange of engrained gender bias and childhood cartoons? I'll leave that to someone else to explore.
----------------
Data Visualization Update
State comparisons:
https://public.tableau.com/views/Coronavirus-ChangeovertimeintheUSA/2_Corona?:display_count=y&:origin=viz_share_link
Rt data
https://public.tableau.com/shared/7FH637YGW?:display_count=y&:origin=viz_share_link
FT data is still the best visualization I have found for country comparisons.
https://ig.ft.com/coronavirus-chart/?areas=usa&areas=gbr&cumulative=0&logScale=1&perMillion=0&values=deaths
A few points to consider:
deaths in the US are falling, but cases are not.
The reopening (in all 50 states now) will impact these numbers, with death lagging new cases, as you would expect.
----------
For the third day in a row, I have found an ongoing discussion around data integrity. This topic is increasingly feeling like a Jerry Springer episode that brings out spreadsheets and illuminates reporting errors instead of surprise paternity results.
Here is a discussion on Twitter by the FT data visualization guy about his concerns on reported data from around the world. Worldometer is, apparently, a problematic data source. There is a lot of insight into data integrity in this discussion.
https://twitter.com/jburnmurdoch/status/1263062077213749250
More on Worldometer's potential issues
https://edition.cnn.com/interactive/2020/05/world/worldometer-coronavirus-mystery/
Here is an article on undercounting causes of death
https://fivethirtyeight.com/features/coronavirus-deaths/
Lastly, there are still unknowns. The COVID Tracking Project is starting to do more subanalyses of racial data.
https://covidtracking.com/race
https://docs.google.com/spreadsheets/u/1/d/e/2PACX-1vTfUQPxkhP_CRcGmnnpUBihnTNZ9Z8pcizII4_sc2o2n3opOoAJdAM4CRTJBI339tou8LWnQrqbTMgH/pubhtml#
There were two articles of note released in Science yesterday. The first article offers further confirmation that coronavirus infection does protect against re-infection. The second article looked at using a DNA-base vaccine, which was able to elicit both humoral (antibody) and cellular (T-cell mediated) immune response. Both articles used rhesus macaques, which are good animal models for humans, but (obviously) not human. The bottom line for both is - good news, but still a lot more we need to understand.
https://science.sciencemag.org/content/early/2020/05/19/science.abc6284
https://science.sciencemag.org/content/early/2020/05/19/science.abc4776
https://twitter.com/GaetanBurgio/status/1263232531568013314
More analysis https://thehill.com/policy/healthcare/498732-two-new-studies-suggest-covid-19-antibodies-provide-immunity
Here is a group of physicians pulling together an impressive review of data on mask-wearing. And it has a clever title - Maskomania
https://www.the-hospitalist.org/hospitalist/article/222464/coronavirus-updates/maskomania-masks-and-covid-19
Here is an article with no direct relation to COVID, but is interesting. Continued data is emerging in support of daily, moderate to low-intensity physical activity. Here is an article about a very large epidemiologic review of those that commute by bike in England spanning many years and 300,000 people. (For full disclosure, many of you know my affinity for my treadmill desk, and my mere selection of this article speaks to my bias for justifying its use.)
https://www.standard.co.uk/news/health/cycling-to-work-can-cut-risk-of-heart-disease-and-cancer-by-a-quarter-a4446131.html
Infographic of the day #1: Brownies. I was aghast to learn that one of our teammates made brownies yesterday without a full appreciation of all the potentially modifiable variables. Despite this amazing oversight, I hear they were still tasty. But that is not the point. I strongly encourage both taste and the satisfaction of knowing you have consciously decided on each brownie attribute! (An entirely intentional brownie that Matt Brill would approve of, I think!)
https://external-preview.redd.it/CHNg0BeQbfKZLJEsqwbp4YhtTm0nzHfLBpKom9QxekQ.jpg?auto=webp&s=667dcaf415d631a00179634314e7d3d4556a1918
Infographic of the day #2:
One realization from college was that I learned about things that allow one to sound erudite, even without a depth of knowledge on a topic. It is a meta-knowledge if you will. A friend sent me this infographic last night that felt like this (he did not make it). When I first looked at it, I thought, "Wow, that is clever." But the more time I spent, the less I think I understood. On further reflection, I think it speaks to the evocative hipness of ven diagrams and dystopian novels rather than some great insight. I am certainly open to feedback if I am missing something.
https://twitter.com/drjonesaa/status/1262107937910251523/photo/1
Bonus Round --- Obscure things edition
A roundup of a few things I discovered recently:
First, serious.
I found an interesting twitter thread about Hilda Hudson, an English mathematician. She established herself as a solid academic mathemetician, despite the cultural limitations placed on women at that time. She explored novel areas of geometry and advanced mathematics, including the underpinnings of theories on disease spread in epidemiology. And she spent time helping design and improve aircraft during World War 1.
https://threadreaderapp.com/thread/1263094231813455872.html
https://www.agnesscott.edu/lriddle/women/hudson.htm
Second, Krampus
I do not celebrate Christmas, but I am fascinated by the variation of the holiday. One such variation from central Europe is the anti-Santa, demon Krampus that punishes the naughty children. (No clarity on how naughty is defined or measured). I am struck by the underlying psychology of trying to scare children into good behavior. Much less, in costume around significant holidays.
https://ridiculouslyinteresting.com/2020/05/20/scary-vintage-krampus-claws/
https://en.wikipedia.org/wiki/Krampus
Clean hands and sharp minds, team
AW
We are starting to see a more constant and robust stream of data around many topics related to COVID. And, as these data are published or shared, each one needs to be thought about critically - what were the data gathering biases? What do the analyses demonstrate? Are the conclusions logical or marred by fallacy?
And, speaking of data, I got a surprising number of responses talking about individual experiences with robot Vacuum cleaners (Dustin' has cousins, it appears). Amongst the various data points, I learned that 2/2 respondents who name their robots name those robots, Rosie. Next line of investigation - why do we apply genders to robots, and why does the Jetson's robot inform the naming of these respondents? A fascinating chain of thought reflecting a mélange of engrained gender bias and childhood cartoons? I'll leave that to someone else to explore.
----------------
Data Visualization Update
State comparisons:
https://public.tableau.com/views/Coronavirus-ChangeovertimeintheUSA/2_Corona?:display_count=y&:origin=viz_share_link
Rt data
https://public.tableau.com/shared/7FH637YGW?:display_count=y&:origin=viz_share_link
FT data is still the best visualization I have found for country comparisons.
https://ig.ft.com/coronavirus-chart/?areas=usa&areas=gbr&cumulative=0&logScale=1&perMillion=0&values=deaths
A few points to consider:
deaths in the US are falling, but cases are not.
The reopening (in all 50 states now) will impact these numbers, with death lagging new cases, as you would expect.
----------
For the third day in a row, I have found an ongoing discussion around data integrity. This topic is increasingly feeling like a Jerry Springer episode that brings out spreadsheets and illuminates reporting errors instead of surprise paternity results.
Here is a discussion on Twitter by the FT data visualization guy about his concerns on reported data from around the world. Worldometer is, apparently, a problematic data source. There is a lot of insight into data integrity in this discussion.
https://twitter.com/jburnmurdoch/status/1263062077213749250
More on Worldometer's potential issues
https://edition.cnn.com/interactive/2020/05/world/worldometer-coronavirus-mystery/
Here is an article on undercounting causes of death
https://fivethirtyeight.com/features/coronavirus-deaths/
Lastly, there are still unknowns. The COVID Tracking Project is starting to do more subanalyses of racial data.
https://covidtracking.com/race
https://docs.google.com/spreadsheets/u/1/d/e/2PACX-1vTfUQPxkhP_CRcGmnnpUBihnTNZ9Z8pcizII4_sc2o2n3opOoAJdAM4CRTJBI339tou8LWnQrqbTMgH/pubhtml#
There were two articles of note released in Science yesterday. The first article offers further confirmation that coronavirus infection does protect against re-infection. The second article looked at using a DNA-base vaccine, which was able to elicit both humoral (antibody) and cellular (T-cell mediated) immune response. Both articles used rhesus macaques, which are good animal models for humans, but (obviously) not human. The bottom line for both is - good news, but still a lot more we need to understand.
https://science.sciencemag.org/content/early/2020/05/19/science.abc6284
https://science.sciencemag.org/content/early/2020/05/19/science.abc4776
https://twitter.com/GaetanBurgio/status/1263232531568013314
More analysis https://thehill.com/policy/healthcare/498732-two-new-studies-suggest-covid-19-antibodies-provide-immunity
Here is a group of physicians pulling together an impressive review of data on mask-wearing. And it has a clever title - Maskomania
https://www.the-hospitalist.org/hospitalist/article/222464/coronavirus-updates/maskomania-masks-and-covid-19
Here is an article with no direct relation to COVID, but is interesting. Continued data is emerging in support of daily, moderate to low-intensity physical activity. Here is an article about a very large epidemiologic review of those that commute by bike in England spanning many years and 300,000 people. (For full disclosure, many of you know my affinity for my treadmill desk, and my mere selection of this article speaks to my bias for justifying its use.)
https://www.standard.co.uk/news/health/cycling-to-work-can-cut-risk-of-heart-disease-and-cancer-by-a-quarter-a4446131.html
Infographic of the day #1: Brownies. I was aghast to learn that one of our teammates made brownies yesterday without a full appreciation of all the potentially modifiable variables. Despite this amazing oversight, I hear they were still tasty. But that is not the point. I strongly encourage both taste and the satisfaction of knowing you have consciously decided on each brownie attribute! (An entirely intentional brownie that Matt Brill would approve of, I think!)
https://external-preview.redd.it/CHNg0BeQbfKZLJEsqwbp4YhtTm0nzHfLBpKom9QxekQ.jpg?auto=webp&s=667dcaf415d631a00179634314e7d3d4556a1918
Infographic of the day #2:
One realization from college was that I learned about things that allow one to sound erudite, even without a depth of knowledge on a topic. It is a meta-knowledge if you will. A friend sent me this infographic last night that felt like this (he did not make it). When I first looked at it, I thought, "Wow, that is clever." But the more time I spent, the less I think I understood. On further reflection, I think it speaks to the evocative hipness of ven diagrams and dystopian novels rather than some great insight. I am certainly open to feedback if I am missing something.
https://twitter.com/drjonesaa/status/1262107937910251523/photo/1
Bonus Round --- Obscure things edition
A roundup of a few things I discovered recently:
First, serious.
I found an interesting twitter thread about Hilda Hudson, an English mathematician. She established herself as a solid academic mathemetician, despite the cultural limitations placed on women at that time. She explored novel areas of geometry and advanced mathematics, including the underpinnings of theories on disease spread in epidemiology. And she spent time helping design and improve aircraft during World War 1.
https://threadreaderapp.com/thread/1263094231813455872.html
https://www.agnesscott.edu/lriddle/women/hudson.htm
Second, Krampus
I do not celebrate Christmas, but I am fascinated by the variation of the holiday. One such variation from central Europe is the anti-Santa, demon Krampus that punishes the naughty children. (No clarity on how naughty is defined or measured). I am struck by the underlying psychology of trying to scare children into good behavior. Much less, in costume around significant holidays.
https://ridiculouslyinteresting.com/2020/05/20/scary-vintage-krampus-claws/
https://en.wikipedia.org/wiki/Krampus
Clean hands and sharp minds, team
AW
Comments
Post a Comment