« first day (2592 days earlier)      last day (2404 days later) » 

1:45 AM
@Glen_b Nice. I am reminded that many classical procedures are proven admissible by showing they are Bayes procedures for some specific prior. This demonstrates that thinking of "implied priors" is not necessarily confined to the Bayesian school.
 
 
2 hours later…
3:34 AM
How to write a CV question.
1. State you don't know any stats. After all everyone else that asks questions must be an expert.
2. Nevertheless be sure to try to cram in as much jargon as possible, because explaining in plain English might make your meaning clear enough to be understood
3. Start with a list of 3 or 4 premises stated with absolute confidence. Don't back down from these premises. After all, you don't know any stats so everything you think you know must be completely true. End with a point of confusion resulting from them, but whatever you do, don't phrase it as a question. Leave it implicit.
4. When people question the basis for your premises don't ask whether they're true - how would that make sense on a site for asking experts about statistics? Instead, justify those premises by trying to bang down a few more premises! We all need more premises.
5. get annoyed when people can't follow what you want. After all, it's perfectly obvious to you what you mean.
....
Am I cynical? Very well then, I am cynical. (I am large, I contain multitudes.)
6
 
 
6 hours later…
9:31 AM
@Glen_b stats.meta.stackexchange.com/questions/3175/… is already our canonical grumbles-turned-into-warnings catharsis. You must have some new material. I identify a Principle of Entitlement: I have a question, so it follows that (someone on) the internet owes me an answer.
It's now common that people in some fields with what looks like a Master's project in something statistical claim to be new to statistics, post a couple of vague paragraphs, and ask to be told (a) how to analyse my data (b) what the code will be. Isn't that the core of the project, outsourced?
 
10:15 AM
@Nick thanks. Yes, perhaps I should ... but will be mostly busy for a few days.
Maybe after that
On the project -- well there's all that data collection, a brief literature survey and writing an introduction. That's got to be worth a Masters, innit?
 

« first day (2592 days earlier)      last day (2404 days later) »