?

Log in

No account? Create an account
color cycle (slow)

Kistaro Windrider, Reptillian Situation Assessor

Unfortunately, I Really Am That Nerdy

Previous Entry Share Next Entry
Mathturbation
a look of abject horror, yikes
kistaro
So I'm trying to get this final paper for Experimental Psychology done. Part of it is background research...

I'm befuddled. How the heck can these people use the F probability function with 22 degrees of freedom when they only used 12 experimental subjects? That's not what I learned in my Stats class, and that explains the shockingly small P value for such a small number of subjects! There's something weird here. And it would be nice if the "Results" section gave all their numbers, even the ones that turned out to have insignificant P... Meehan, Triggs, and Phillips would never have passed Dr. Holt's class, no no no.

  • 1
I'm afraid I can't comment on the stats.. how do you find Experimental Psychology, though? It's an option for me next year:

http://www.cam.ac.uk/admissions/undergraduate/courses/natsci/part1b.html

Experimental Psychology was quite a bit of work, but very interesitng and I learned quite a lot. If your course is taught like mine, only the last 1/3 of the course will be you performing experiments. Fully half of it will be training you in the proper format and style for writing up scientific papers, valuable in any scientific field.

During the course, I performed a number of experiments as a participant; Dr. Holt had the entire class tabulate the results, draw conclusions, and then write up a particular part of it. It's the last we're graded on, although there's a lot to learn form the whole thing. The last part was the really interesting bit in which we designed and performed our own experiments- now I have to write up a complete experiment paper by Monday afternoon. (Egad.)

So in conclusion- it's really interesting and really fun for me, but there's quite a lot of work involved- and it helps to find it interesting.

Your college should have subscriptions to online academic journal publications. Track them down and read a few psych research papers, preferably on topics that interest you. Was it interesting? Can you see yourself spending part of Finals Week writing one? Can you see yourself inventing and then performing such an experiment? Draw your own conclusions...

Computer programming skill is a big help with this, incidentally. Most experiments are computer-based; I got bonus points for doing an experiment that was computer based without being one of the standard implementations of an implicit attitude test. It did take quite a bit of work to set up, however... just remember that the class probably will take quite a lot of your time.

Just don't sue me if I'm wrong, I expect Cambridge University to differ from Washington University quite a bit...

Thanks for your comments! The courses that I have so far are quite different to that which you describe (and I take just four courses over the whole first year, three of which are Natural Sciences and one mathematical course, though still within the NatSci Tripos - not the same courses the mathematicians take). I've not learnt anything about writing papers so far, but I expect that will have to come at some point!

The course content does sound quite interesting, indeed - and anything where you get to design your own experiments must be! I should have a look at journals - though of course we do have subscriptions to most journals, some of them require ATHENS authentication which I neglected to pick up, while others want you to be on the .cam.ac.uk domain, which I am obviously not at school. If they authenticate via Raven then I can access them.

Again, thanks for your thoughts - I don't decide until later next year, of course, but we'll probably get asked for our preliminary choices towards the end of next term.

No problem! To hit Washington University subscriptions, I have to be within Washington University's IP-space, but I'm still getting to the subscriptions from home- I'm using Remote Desktop to connect to the CEC's central server for the purpose for students who need off-campus access to CEC resources, and that can hit the system just fine. Do you have a way to Remote Desktop into your computer labs?

Most of our internal resources are Raven controlled, so for a lot of users that is sufficient:
http://raven.cam.ac.uk/

The only authorised way I know of to be within the university's IP range is to use their VPDN service:
http://www.cam.ac.uk/cs/netdiv/vpdn.html

I might have to apply for that, for accessing journals when away from the university. The setup they describe routes everything through the CUDN while connected, which is a bit of a pain. (I'm not sure what the 'undue complexity' is with the Windows 2000 client if the XP one is fine, they appear to be very similar to me.)

I haven't taken stats proper, but I've done a lot of stat-like stuff in Ecology this semester, and I really have to wonder how the hell you get 22 degrees of freedom from 12 subjects.

Well, besides taren_'s wonderful equation.

*Shakes a fist at you*

Curse you for drawing me in with your carefully worded title and slapping me in the face with math!

Wings can form fists? O_o

Hey, you got exactly what the title said. It's a term that one of my professors, Dr. Robert Pless, uses frequently to describe "mathematics that doesn't really get anything done and serves no purpose except, perhaps, for great pleasure by whoever is doing it. Furthermore, most people do it at some point in their lives, then lie about it."

Dr. Pless is an amusing professor.

  • 1