Embedded Files
Resources for Bayesian Modeling and Inference
Michael Jordan Course: https://people.eecs.berkeley.edu/~jordan/courses/260-spring10/lectures/index.html
Resources for Bayesian Decision Theory:
Resources for Causal Inference I found useful:
Peng Ding's personal website- has links to Fan Li, Guido Imbens and Stefan Wager's lectures.
Peng Ding's First Course in Causal Inference
Bayesian Causal Inference: A critical review by Fan Li, Peng Ding, Fabrizia Mealli
Online Causal Inference Seminar
Resources for Stat Inference (especially for undergrad courses/early grad courses):
Matthew Stephens' website (especially for undergrad courses/early grad courses)
Steven Goodman's the p-value fallacy and the Bayes Factor
Resources for good presentations:
Andy Stein's website mentions a checklist for presentations, and an evaluation form for a presentation. Both can help prepare you for one.
Resources for manuscripts:
This article by Jacob Bien gives a checklist for your next article submission including capitalization, logical flow, and citation. For example, he writes `Use \citep{} for a parenthetical and \citet{} for a noun. Examples: (a) “The lasso \citep{tibs96} uses an L1 penalty.” (b) ”\citet{tibs96} shows that...” One places a citation in parentheses as in (a) when you could read the sentence without the citation and it would still'.
Other Interesting links:
Read: Multiple Testing Joseph P. Romano, Azeem M. Shaikh, and Michael Wolf,
Frank Bretz et al: Graphical approaches for multiple comparison procedures using weighted Bonferroni, Simes, or parametric tests (slides)
Burman et al. A recycling framework for the construction of Bonferroni-based multiple tests
Bretz et al: Advanced multiplicity adjustment methods in clinical trials
FDA Sirisha Mushti slides on multiplicity adjustment
Tipping Point Analysis: as a choice for sentisitivity analysis in clinical trials
Mirat Shah et al The Drug-Dosing Conundrum in Oncology - When Less Is More
A Story on Confounding (via Dr Arnab Chakraborty, ISI):
A fortune-teller once claimed he could predict how long a person would live just by measuring their palm. Apparently, all he needed was the length and width of your hand, and voilà—your life span was revealed! The scientific community was left in a daze, with human biologists scrambling to figure out how they missed the "Handbook of Life" in their studies!
That's when a famous statistician, Mr. Ronald Dubin (disclaimer: names don't represent real people) enters the picture. He collects information about hand length (a), hand-width (b), binarized gender, and age at death of 500 people.
When he plots the ratio of hand length to hand width (a/b) to the age at death, this is what he sees:
But once he colours each data point by gender, he sees:
In fact, after correcting for gender, there is no apparent correlation between the hand dimensions and age at death.
Moral(s) of the story:
Females had longer hand length to width ratio in the study
Females were found to live longer
And finally, the dimensions of your hand cannot help predict how long you will live! ( at least this study can't help you infer that ;) )
Page updated
Google Sites
Report abuse