PSA

In the word "spurious," just in case you were in doubt, there is a "y" sound between the "sp" and the "ur."

Everyone who has said it aloud in my stats class, including the instructor, says "spur" + "ious," as if it ought to mean "full of spurs." Roy Rogers takes on multiple regression. Or something like that.

Perhaps I am sensitive to missing "y" sounds because I spent 23 years telling people my last name had a hidden "y" sound in the middle. My father-in-law still doesn't know how to pronounce it. It's not that weird, I'm telling you.

When I was first contemplating a return to grad school, I thought it would be great to have a dashing and handsome math geek for a husband because he could help me with my stats questions. Alas, it doesn't work like that. If I am wrestling with something he finds trivially easy, it is not a recipe for a helpful conversation. So I will tell you, my pals in the computer, that I am utterly flummoxed by the chapter on curvilinear regression that I set out to read at 2pm and still have not finished.

Random example: "The cross-product vectors represent the interaction. The approach is similar when the design includes continuous and categorical independent vectors. The latter are coded in the usual manner and the former are coded by orthogonal polynomial coefficients." [faint whimper]

None of the stuff we have been doing is really very complicated. Simple linear regression looks at the way that one variable can predict or explain another, like SATs and college GPA. Multiple regression throws other explanatory variables in there: motivation bumps GPA up and weekly alcohol consumption brings it down, and we can figure out how important those effects will be. Curvilinear regression is just the idea that lots of important relationships don't look like straight lines and we need to be able to explain them too. So I went sailing into the chapter with confidence, but there is no wind in my sails at the moment.

Probably I will get more out of tomorrow's lecture if I go to bed than if I stay up blogging, though, you figure? Over and out from a weary moi.

Comments

[Tracy]

I am whimpering just looking at the words curvilinear regression. I will now cope by ignoring them completely since I do not have to learn them at all :-).

I am more upset that I am probably saying your last name wrong. I will now politely request (I was going to say demand and decided that was a bit too strident) that you drop everything and e-mail me to let me know where the hidden "y" sound should be, so I can learn to pronounce it correctly. You know, after you get all of the important stuff done - just put it on the to-do list "e-mail Tracy with correction". Thanks!

[mary]

Really? There's a y sound in spurious? Is this like duty -- some say 'dyouty' and some say 'dooty'?

I learned something here today. Not stats, but something just the same. :)

[giddy]

What about coupon? I say spyurious, but on the other hand I say coopon rather than kyoopon. I suspect spurious without the Y sound is an accepted variant--or will be soon if it isn't already.

I have a hidden Y sound in my first name, so I feel your pain, maybe even more since my hidden Y sound didn't go away when I got married.

Good luck with curvilinear regression; sounds like a nightmare!

[CJ]

I vote for coo-pon, though I grew up saying cue-pon. I can't really justify cue-pon on etymological grounds. I'm not much of a prescriptivist, though. :-)

[Maria]

Never heard of spYurious in my life. I'm not as educated as you are ("curvilinear regression"? I'm feeling faint!) though, so maybe I just missed it.