recentpopularlog in


« earlier   
"Reproducibility — Replicability: P-values and the Larger Questions," Andreas Buja
"Statistical Thinking: Statistics = “quantitative epistemology”
Statistics = the science that creates protocols for the acquisition of
qualified knowledge."

[Yes! I like Buja's definition of statistics, as well as his framing of the problems we face and possible solutions. Journals should accept well-planned studies, blind to what the p-values turn out to be. Stats methodologists should develop more post-selection inference, graphical inference, etc. so that we can honestly report our whole data analysis process and not just the one branch that led to a significant p-value.]
statistics  epistemology 
7 hours ago by civilstat
Maps are a serious fake news threat, and bots may be making them now
A new analysis of “viral maps” published in Cartography and Geographic Information Science examines the more than 500 maps that sprang from Silver’s original one, using Google’s Cloud Vision image analysis platform. The study also points out maps that were created with the pure purpose of misinforming. One egregious example, which claimed to represent an electoral map if only taxpayers voted, was simply a retitled map of something else entirely.
According to Anthony Robinson, an assistant professor of geography at Penn State University who conducted the study, maps are a particularly ripe format for spreading misinformation on the internet because we’re so used to trusting them as fact....

Where there’s misinformation, there’s usually bots involved. Robinson thinks that right now, maps like the one claiming to show what it might look like if only taxpayers voted are likely being disseminated and amplified using bots....

But it may be only a matter of time before bots begin to generate maps automatically. The technology exists: Robinson points to the increasing prevalence of auto-generated video called deepfakes...

While Robinson thinks media literacy is a laudable goal, he doesn’t believe it’s practical. Instead, he’s hoping to use machine learning image detection algorithms like those from Google Cloud Vision to trace the provenance of maps and how they spread online. Because algorithms have the capability to find both exact and partial images matches, they could potentially show users the lineage of a map’s journey through cyberspace.
mapping  cartography  misinformation  epistemology  bots 
7 days ago by shannon_mattern
Victorian Doctors Didn't Treat Women With Orgasms, Say Historians - The Atlantic
“One of the big takeaways for me is that the peer-reviewed process is flawed. Peer review is no substitute for fact-checking,” she added. “We need to fix this, and we need to start checking other people’s work, especially in history.”
methodology  epistemology 
11 days ago by shannon_mattern
Why it’s as hard to escape an echo chamber as it is to flee a cult | Aeon Essays
- difference between epistemic chamber and echo chamber - how people in an echo chamber are not necessarily irrational but is like an addiction This is a good essay on understanding information networks, how belief systems form, difference between an epistemic chamber and an echo chamber (and how to escape them)
information  epistemology 
12 days ago by k4rtik

Copy this bookmark:

to read