A commenter wrote (by email):
I’ve noticed that you’ve quit approving my comments on your blog. I hope I didn’t anger you in some way or write something you felt was inappropriate.
My reply:
I have not been unapproving any comments. If you have comments that have not appeared, they have probably been going into the spam filter. I get literally thousands of spam comments a day and so anything that hits the spam filter is gone forever. I think there is a way to register as a commenter; that could help.
You've not posted comments by me, also (pretty innocuous ones expressing disbelief about the power
of certain models regarding increased partisanship
in Congressional elections–all right, maybe they were intellectually challenging to you but they weren't personally insulting). It's fun to play devil's advocate without taking the risk of retaliation (of grants being disapproved, papers rejected, etc)–this is why reviewing is anonymous.
Numeric:
I'm not sure if you're joking, but just in case you're not, let me repeat, this time in all caps for emphasis:
I APPROVE ALL NON-SPAM COMMENTS THAT COME IN. BUT I GET THOUSANDS OF SPAM COMMENTS A DAY. IF YOU EVER ENTER A COMMENT THAT IS NOT POSTED, IT GOT TRAPPED BY THE SPAM FILTER. I CAN'T CHECK THE SPAM FILTER–I GET THOUSANDS OF SPAM COMMENTS A DAY.
Hmm, maybe I should say this again; I don't think the message got through.
IF YOUR COMMENT DID NOT APPEAR, IT GOT TRAPPED IN THE SPAM FILTER. IT DOES NOT MEAN THAT I "DID NOT APPROVE" YOUR COMMENT. IT GOT TRAPPED IN THE SPAM FILTER. TRAPPED TRAPPED TRAPPED.
I had the same problem for a while.
So correct me if I'm wrong but what I think your trying to say is that if a comment did appear, it got trapped in the spam filter?
Andrew: "I think there is a way to register as a commenter"
It's even easier than that. There's a link above the comment box that says "Sign in".
Click this link and you have a variety of options for signing in. I use an old Live Journal account, which is why there's a little pencil next to my comments. Andrew's comments have a scissors next to them, which is a tiny logo for another company.
The advantage to this is that your comments appear right away and Gelman doesn't have to approve them, saving him time best used elsewhere.
I wonder if there wouldn't be some trick (even easier than those annoying Captcha's) to separate the human produced gems from the machine produced spamvertisements.
I would still like to know about the 3 distinct modes of statistical inference that apply when someone has been able to survey an entire population (1200). Can someone tell me what I can infer with 370 reponses to a survey that was mailed to all 1200 members of the population?