T-Distribution...wha'?

In honor of my Math Geek bf who passed his Actuarial Exam today (yay to him!), I decided to hunt down something math related on Etsy. I found something that was rather cute, though I don' t remember learning about it in school (me being an English major tries to ignore math formulas like the swine flu).

I present to you the T-Distribution plushie:According to the seller's description:

"The normal distribution comes in really handy when computing confidence intervals and performing hypothesis test, when you have a large enough sample. But when your sample is too small, you'll get better results by using the t-distribution. "

Right. Because I understood that. Regardless, the plushie is cute and comes with it's very own formula in the back:


Yup, makes total sense to me. Anyhow, if you have a math geek in your life, you can get this for him/her for a total of $8.00. That's quite a bargain, if you're willing to take an hour long explanation after giving them this gift on just what a t-distribution is.

Comments

wat said…
That's totally NOT a df = 200 t-distribution. 200,000, maybe.
Ookami Snow said…
Full disclosure: My wife is the person that makes these.

@wat - The patterns that she uses are generated from the statistical program R, so the shape of the three lines is (give or take some error from sewing) exactly the shape of DF = 1, 2, and 200.

Mathematically as there are more and more dfs the t distribution gets closer and closer to the z distribution. And in general once there is 30 degrees of freedom it is "close enough" to the z. So a t distribution with df=200 will essentially be the same as a t with df=200,000 which will essentially be the z distribution. (Like how the number 9.999 is very close to the number 9.99999999 which both are basically 10.)

Crap, is my statistics nerdiness showing through? ^_^

@Suze - In any case, thanks for the review on your blog! Nicole gets a real kick when somebody else enjoys the geeky creations that she makes.
wat said…
This comment has been removed by the author.
wat said…
@Ookami - wow. I'm impressed that she took the time to map out the t-distribution in some modeling software. (And am loving the fact that she's got stuff like lognormal plushies as well.)

I suppose I should've qualified my comment by saying I'm purposely obtuse 90% of the time. (The other 10%, that's on accident.) I got my BS in Math and do appreciate the shape of the plushie. Although, I must've missed the class mentioning how "quickly" the t-distribution converges to the normal curve based on degrees of freedom.
Suzy said…
This comment has been removed by the author.
Suze said…
My pleasure! I'm not a big math geek, more of a literary geek, but it was adorable nonetheless. :)