- American Geophysical Union honors Peter Gleick.
- What does using SPSS say about you? (Via Marginal Revolution). I disagree that R users “do not care about aesthetics.” R can make much nicer graphs than other packages.
- Does the Daily Mail website get 100 million unique visitors per month? If so, did Michael Jackson sell one billion records?
- IRB difficulties, social science division
- Unsafe injections: “This can’t happen in the United States, this is a Third World thing.”
- What medication was Lanza on? “It may well turn out that knowing what kinds of guns he used isn’t nearly as important as what kind of drugs he used.”
- Aaron Swartz in his own words.
Thanks to Patrick Vlaskovits.
“When it’s extremely low cost to perform inference, you are likely to perform a lot of inferences. When your first regression gives a non-result, you run a second one, and a third one, etc. This leads untrained researchers to run into multiple comparisons problems and increases the risk of Type I errors.” But it also enormously improves the chances of finding something publishable.
Seth: “Untrained researchers”– ha ha! It’s better to find something than nothing even if you can’t put a p value on the something. In the early days of exploratory data analysis respected statisticians criticized it because they said it would lead to finding patterns where no patterns existed.
Peter Gleick
Seth: Thanks, fixed.
They make grade school kids fill out human subject consent forms for science fair projects now:
https://www.sciencebuddies.org/science-fair-projects/project_src_safety_human_subjects.shtml