Paul Cleary
@paulcleary.net
Epidemiologist and data/IT nerd, working in infectious disease surveillance in the UK and abroad. Has a sort of blog here: https://prcleary.github.io/#. Understands R and Python, French and German. Interested in too many things.
in my defence, I was left unsupervised
August 30, 2025 at 12:27 PM
in my defence, I was left unsupervised
Is Indonesia the nicest place in the world? I just came back to my room to find the housekeeper has decided I need a better bookmark than a torn rail ticket
February 6, 2025 at 1:34 PM
Is Indonesia the nicest place in the world? I just came back to my room to find the housekeeper has decided I need a better bookmark than a torn rail ticket
just had a surreal exchange over an eBay refund - can any linguistician explain this to me? #linguistics
January 17, 2025 at 12:28 PM
just had a surreal exchange over an eBay refund - can any linguistician explain this to me? #linguistics
"Univariable" is not a word
December 19, 2024 at 12:00 AM
"Univariable" is not a word
thanks for the excellent suggestions
December 5, 2024 at 9:49 AM
thanks for the excellent suggestions
Should public health training (or demonstration of equivalency) be required to progress to the most senior public health epidemiology roles? Discuss.
December 4, 2024 at 10:38 AM
Should public health training (or demonstration of equivalency) be required to progress to the most senior public health epidemiology roles? Discuss.
first draft of my 2025 reading list is looking a bit male, pale and stale - any suggestions? for literature, I tend to go for dead authors who we are still talking about (sorry living authors), but I am open to persuasion
#booksky
#booksky
December 1, 2024 at 11:36 AM
first draft of my 2025 reading list is looking a bit male, pale and stale - any suggestions? for literature, I tend to go for dead authors who we are still talking about (sorry living authors), but I am open to persuasion
#booksky
#booksky
It's been a while, but I seem to remember my lecturers saying that it is a rate if the denominator has different units to the numerator, as in e.g. cases/person-years, rather than being about time as such - but in practice it is used loosely like many epi terms
![A risk is a proportion, that is, a measure in which the denominator includes the numerator. For example, the risk of developing lung cancer is the proportion of a group of people at risk (denominator) who newly develop lung cancer (numerator) over a specified period of time (e.g., the risk of lung cancer can be 10% over 20 years in heavy smokers).
A rate is a measure of change in one quantity per unit of another quantity. In epidemiology, a rate is often used as synonym for incidence rate, which is the change in risk per unit of time. For example, a risk of 10% over 20 years can, if constant, be expressed as a rate of 0.5 per 100 and per year [rate = risk + time = 10% + 20 years = 0.5% per year].](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:piubidfrubqokdw2nvla7cvv/bafkreiackzfomg5vtwm6lpxqlsoakt4nvckfb64gryjywbjzs73dk7zu4e@jpeg)
November 26, 2024 at 9:18 PM
It's been a while, but I seem to remember my lecturers saying that it is a rate if the denominator has different units to the numerator, as in e.g. cases/person-years, rather than being about time as such - but in practice it is used loosely like many epi terms
Anyone else finding that their DAGs look nothing like the neat examples in the textbooks? #causality
November 24, 2024 at 11:40 AM
Anyone else finding that their DAGs look nothing like the neat examples in the textbooks? #causality
I was under that that day - glad to hear it is improving
November 19, 2024 at 8:27 AM
I was under that that day - glad to hear it is improving