Limits of Inference

Home
Archive
About
Overdiagnosis: why scientists and statisticians think we should test fewer people
A targeted testing strategy is the gold standard when screening for disease. This is the best practice for a reason.
Overdiagnosis: false-negatives and sensitivity of RT-PCR [1]
why false negatives aren't as much of a problem in practice
Overdiagnosis: technical versus clinical false positive [2]
Whether or not a false positive is an error can depend on who you ask
Overdiagnosis: how early detection changes the definition of disease itself [3]
what does it mean when disease incidence is uncorrelated with mortality. can you have cancer even if you will never be sick?

December 2020

The best and worst of public health
being healthy means saying yes to life
1

November 2020

Understanding Risk From Viral Disease
Infection is a risk factor, not a disease itself

October 2020

Science does not solve our problems. People do.
when we look only to scientists for solutions, we will come up short
Defining Success: Zero-risk is not an option
We still do have to learn to live with all viruses

September 2020

What is a probabilistic value proposition?
Imagine if Gmail delivered 95% of emails to the correct person’s inbox and sent the other 5% to a different person’s inbox.
Theory over Data
how scientific reasoning and logic failed during a public health crisis

May 2020

Debunking the myth of purely ‘data-driven’ decisions
data is a great tool, some expect too much of it

April 2020

What is inductive reasoning?
Why do we put so much faith in a fallible process?
© 2024 Clare Gollnick
Privacy ∙ Terms ∙ Collection notice
Start WritingGet the app
Substack is the home for great culture