November 14, 2004

Stats update - I passed!

Posted by Maria

I forgot to post a follow-up on my recent struggles with statistics. Well, it may be that blogging on something as opposed to actually doing it is shockingly useless displacement activity, but that doesn’t mean it’s completely wasted time. By the time I got to the exam, all aggravation had been expressed and I had some great suggestions for books to use if I didn’t scrape the pass this time round. Plus, I’d ditched the parts of the course I was just not absorbing (poisson distribution and other topics whose very names I’ve since erased from memory ) and practiced using the parts I understood.

Luck played a part again (is this statistically significant, 3rd time in a row?). The really hard trick question, which I considered for 5 minutes and then skipped in favour of a straightforward regression, turned out to have been so fiendishly difficult that no one answered it, and it was excluded from the marking. My performance probably said as much about a well-practiced ability to do exams - by scouring the paper for every last point - as it does about my statistics ability. As I left the exam, I figured that with a bit of luck I’d score 12 out of 20. Which I did.

So, what have I learnt? A good bit more about the basics than on my two previous passes. This time round, I actually understood and used regression, and couldn’t understand why it had seemed so impossible last time. And a little bit more about probability, but a very useful little bit. I do agree with Kieran and other commenters that statistics is a really cool and useful tool, and I wish I could wield it better. But now I have (another) piece of paper that says I can do that passably well.

Thanks everyone for the advice, sympathy, and above all the book recommendations. I expect they’ll be useful to many struggling with stats as the pedagogy catches up (or slows down) to non-mathematical students’ needs.

Posted on November 14, 2004 09:37 PM UTC
Comments

Hurray! Well done.

Cf jokes like “What do you call the person who graduated last from his class in medical school?”, etc.

It seems like the exam-taking skills inculcated in the preparation for the Inter and honed for the Leaving are still in place. :)

Posted by Kieran Healy · November 14, 2004 09:54 PM

Well done! I wonder if exam-taking skills are ever lost. I doubt it, but I have no desire to find out the hard way.

Posted by John Quiggin · November 14, 2004 10:31 PM

Yeah, I’m with John: these days I prefer to hone my exam-giving skills.

Posted by Kieran Healy · November 14, 2004 11:17 PM

I’m one who always seemed to manage to throw myself into the endzone for that last second touchdown when it came to “getting” stats (and passing). Always managed to figure it out in the end, but I fortunately managed to avoid ever having to actually teach econometrics. Understand too well why it’s “hard” for students and understood not well enough to find a better way of teaching it….

Posted by Atrios · November 15, 2004 12:20 AM

Yay, congrats! So now that this is out of the way, will we be seeing more of you around here?:)

Posted by eszter · November 15, 2004 12:53 AM

Well done, and remember that some subjects really are so counter-intuitive that you can’t just guess the answers. I seem to remember that in Italy they accept that engineering students will fail a year and are all the better for it.

I hope you don’t mind that I pass on the link to this to a university administrator friend.

The important thing, however, is that you join the club who (can) interpret complex stats based arguments. (The Lancet study for example.)

To second Ezster, let’s see more of you now!

Posted by Backword Dave · November 15, 2004 01:25 AM

Thanks a mill! Yes, I really have no excuses now…

Posted by Maria · November 15, 2004 03:12 AM

Mind you, unlike the New World, most of the Old World’s architecture is living proof that you can build huge and beautiful things with historical knowledge based on observation, a plumb bob and a good eye.

So the Italians have something to encourage them in the belief that failure in engineering ain’t necessarily the end of the world.

Posted by David Tiley · November 15, 2004 03:38 AM

Could you will some knowledge my brain?

I’m enrolled in a stats class next semester (through the math dept., not my poli sci dept.) and have not opened a math text book since high school (almost 6 years now).

Posted by Gozer · November 15, 2004 07:36 AM

That should be “to my brain”.

As you can see I’m ill equipped to take stats. :)

Posted by Gozer · November 15, 2004 07:39 AM

Well done!

Posted by Chris Bertram · November 15, 2004 09:11 AM

I asked this earlier on the Lancet thread, where it was off topic and properly drew no responses — possibly someone here will have an idea. Is anyone aware of a respectable university-level stats course or sequence of courses avalable online that they can recommend?

I don’t have the time available to actually show up for a college statistics course in person, but I’ve become desperately annoyed with the fact that I haven’t got the knowledge to evaluate controversies like the one over the Lancet article on any basis other than that I generally trust the people making one set of arguments.

Posted by LizardBreath · November 15, 2004 03:05 PM

You can check out MIT’s Open Course Ware project at http://ocw.mit.edu/. It’s not like being in the class, but they have lecture notes and so on up there. The quality varies from class to class and professor to professor, naturally.

Posted by anand sarwate · November 15, 2004 07:33 PM

The comments in parathesis after “luck played a part” was brilliantly funny. You certainly deserved a passing grade - just based on that!

Posted by pgl · November 15, 2004 07:51 PM

Lizardbreath,

You might check out the material for the Society of Actuaries Exams 1, 3, and 4. The study guides are quite good and the recommended materials tend to be designed for self-study.

If you want more info, check out actuary.ca.

Posted by SamChevre · November 15, 2004 08:30 PM

Nicely done.

Posted by Greg · November 17, 2004 02:33 AM
Followups

This discussion has been closed. Thanks to everyone who contributed.