EPIGRAM: Helen! you finally made it!!

Post Epigrams

We should post the epigrams as they currently are drafted. If anyone has written a paragraph to explain/explore one, it would be great if you would post those here.Thanks!

Hey All Epigrammarians!The romantic epidemiologist has entered a list of epigrams on the right side of the page here. They need some editing (my bad handwriting filtered through Dr. Kim's creative English), but the basic idea is that we have agreed to write short explanations of each one of these. SHORT is the key - not more than 100 words. Use New Post to do this. In the Title box, put the epigram. Then after you post it, our blogmeister will link your post to the epigram on the right. Then after that, anyone can add a comment to your paragraph.OK?

EpiGrams

Correlation does not imply causation
Finding nothing is finding something
*All models are wrong, some models are really wrong
*Choosing the wrong model creates bias
*Overlapping CI do not mean there's no difference
*Before chossing study design be sure you know the hypothesis
*There is generally a tradeoff between Type I and Type II errors
*Absence of evidence is not evidence of absence
*Non-differential exposure misclassification biases towards the null (usually)
*Avoid bias, control confounding
*All models are wrong but some are useful
*No testing for counfounding
*If it doesn't vary it can't confound

EPIGRAM

Post Epigrams

Monday, April 28, 2008

Helen! you finally made it!!

No comments:

EpiGram?

EpiGrams

EpiGram Meeting

EpiGrammarians

Blog Archive