Tuesday, July 10, 2012

Summer Questions

Been a while since I hung out the shingle -- so here's a place for your questions/comments/suggestions that don't fit with any particular post.

On a different matter, I've noticed that my older posts don't get comments.  This strikes me as odd because I don't close comments.  And recent comments are always shown at the bottom of the page, along with a subscription to comments rss feed on the right hand side.  If they weren't being read, no surprise that there are no comments.  But some posts have had most of their reads months and even years after the original appearance.  Any ideas?


Evan Haliburton said...

Comment sections usually go dead within a few days (a week or two for slow-moving sites, a few hours to a day for high-traffic megablogs) because people prefer to engage when they think they're part of a conversation. Oddly enough, it seems people actually *don't* like getting the last word.

VigiliusH said...

Thank you so much for your excellent blog. I come here mostly for the links, but I read very closely everything you say. I live two feet below sea level here in Gretna, Louisiana and am mostly curious about whether the Harvey Canal control structure will allow me to continue to live here for another ten or twenty years after which I expect to be dead. Anyway, I mostly come here for your time-sensitive blogroll which allows me to follow blogs which do not update daily and it saves me the effort of creating my own feed. So, thanks very much, you seem to be a real working scientist and not a full-time blogger and I appreciate very much the effort you spend on this blog.

dbostrom said...

I've often thought that a blog post should "bubble up" automatically-- be promoted-- when that post earns new comments. Some posts require a lot of chewing and digestion while others are more ephemeral and quickly swallowed, yet they all go in a fixed stratigraphy.

Hangover from original intention of "blogging" software: chronological narrative of everything regardless of importance or triviality.

Robert Grumbine said...

Evan, dbostrom: Sounds good. I'll have to keep in mind the short window for responses. I'm ok with a very slow conversation.

Thanks Vigilius. I am indeed a working scientist, so all blogging is in addition to my day job. Please do suggest topics you're interested in as well.

VigiliusH said...


Well, I am really just a member of the audience here but I guess I do have a bit of a question.

The opinions re climate change I currently have are borne out of a sense of the relative plausibility of competing stories rather than of any genuine knowledge of my own. Some things I can follow, such as your classic post of July 26, 2011 "How not to compute trends." OTOH I when I visit at tamino's I usually leave feeling like a bear of very little brain. I take the little bit of process control statistics that I really know and then I go off looking up Poisson distribution or whatever on wiki, trying to follow the discussion, often with little luck.

So finally getting to my question, could you recommend some web resource I could devote, say, 10 personal study hours to, that would elevate my grasp of statistics enough to follow these things better? (The one thing I know for sure about this is that the formulae always look a great deal more complicated than they really are once you know what they are getting at.)

dbostrom said...

If Dr. Grumbine doesn't mind, let me pipe up and recommend to VigiliusH the statistical explanations at Science of Doom.

(list of explanations appears in reverse order at that link)

The explanations handle statistical basics that other bears of very little brain (me, fer instance) can follow and they're created so as to be relevant to climate behavior.

I'd suggest reading SoD's pedagogy before delving into the comment threads there; there's an awful lot of noise, as usual.

Robert Grumbine said...

Mind? I encourage it! I may read a lot (well, absolutely do), but there's even more out there to read. So I miss a lot.

For this question, I'll encourage reader contributions even more than usual. I realized in thinking about answers that I still get most of this sort of information from books rather than the net. And I'll join the two of you in finding some of tamino's statistics hard to follow. Not that I wouldn't get the glorious details of Laplace's generating functions for statistical distributions if I sat down for a while with a good book. But that it's going to take more than a blog post for me to get it.

For statistics in climate, two books come to mind:
Statistical Analysis in Climate Research, by Hans von Storch and Francis Zwiers
Physics of Climate, Piexoto and Oort

Though the latter carries the title of 'physics', the authors are observationalists and spend a fair amount of time on statistics needed to go from the few observations to a full description of climate. Drawback to the von Storch and Zwiers is that it is pricy.

Online, Nick Stokes, http://moyhu.blogspot.com goes in to some details, particularly as apply to constructing global surface temperature analyses. (I see I have to add him to the blogroll. Thought he was already there.)

Two books on statistics that I like a lot, don't require any math background to speak of (4th graders should be ok) are inexpensive, and get you quite a long way with fundamental concepts of probability and statistics are:
How to Lie with Statistics, by Darrell Huff
Lady Luck, by Weaver

Aside: I almost never use the 'Dr' title. In conversation, even a slow conversation like blogs, I'm Bob.

VigiliusH said...


Okay, I have my homework assignment. Actual books! Maybe this will help me with my internet addiction. OTOH my congresscritter is a guy named Scalise so I'm not sure mow much good it will do me to be a better-informed citizen.