3 ms·

Can we discuss the impact of visualization parameters on perception? On first look it seems Europe is way worse than China regarding case numbers. But actually

by pintxo 7y ago

Can we discuss the impact of visualization parameters on perception? On first look it seems Europe is way worse than China regarding case numbers. But actually checking the scale for the bubbles shows that they are categorical and the largest bucket is ">12,000 cases" thereby putting China (80k) and Italy (12k) into the same visual bucket. Why would you do this?

GBiT 7y ago
Yes. I use the different map - https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6 https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.h...
hef19898 7y ago
Good question. At least the bubbles aren't red...
squiggleblaz 7y ago
Two possibilities: - the goal isn't to convey information, it's to appear to convey information so they can defend themselves later on - we tend to hire artists who are concerned with appearance into roles that require visual engineers The former seems to be something that the UN struggles with. For instance, they can't indicate how many people have coronavirus in Taiwan because they need to consider the feelings of their member states. And they can call one recent coronavirus "Middle East respiratory syndrome", whereas another has to be called "SARS-COV-2" to avoid racist implications of a geographic name. The latter is a widespread problem. For instance, once upon a time, software design gave consideration to making it easy to approach and use, especially by people who might be using computers for the first time. Nowadays, it's primarily about making sure the software looks fashionable. My trust for international organisations has gone down the loo thanks to this outbreak. I am sad.
- sails 7y ago
  This article has some decent thoughts on the quality of data-viz, which has been pretty shocking [1]. ArcGis seems to have pretty bad scaling generally https://www.futurehealth.live/blog/2020/2/10/coronavirus-dataviz-sucks https://www.futurehealth.live/blog/2020/2/10/coronavirus-dat...

by pintxo 7y ago

GBiT 7y ago

Yes. I use the different map - https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6 https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.h...

hef19898 7y ago

Good question. At least the bubbles aren't red...

squiggleblaz 7y ago

Two possibilities: - the goal isn't to convey information, it's to appear to convey information so they can defend themselves later on - we tend to hire artists who are concerned with appearance into roles that require visual engineers The former seems to be something that the UN struggles with. For instance, they can't indicate how many people have coronavirus in Taiwan because they need to consider the feelings of their member states. And they can call one recent coronavirus "Middle East respiratory syndrome", whereas another has to be called "SARS-COV-2" to avoid racist implications of a geographic name. The latter is a widespread problem. For instance, once upon a time, software design gave consideration to making it easy to approach and use, especially by people who might be using computers for the first time. Nowadays, it's primarily about making sure the software looks fashionable. My trust for international organisations has gone down the loo thanks to this outbreak. I am sad.

sails 7y ago
This article has some decent thoughts on the quality of data-viz, which has been pretty shocking [1]. ArcGis seems to have pretty bad scaling generally https://www.futurehealth.live/blog/2020/2/10/coronavirus-dataviz-sucks https://www.futurehealth.live/blog/2020/2/10/coronavirus-dat...