Emilee Rader » data

results!

emilee — Sun, 17 May 2009 06:13:59 +0000

after about a month of on-and-off panic about my dissertation experiment analysis (and too many marathon sessions with crappy R documentation), i now feel confident in saying that YES, i do have some pretty interesting results! i’ll be working on writing everything up for a rapidly approaching paper deadline; i’m guessing i’ll be posting bits as i work on figuring out how to say what needs to be said in the results section, and how best to describe what i think it all means.

i also have to say, what the heck did people do before the internet? i mean, here i am using open-source statistical software, doing fairly nonstandard analyses for my field. and yet, it seems like no matter what problem i encountered either with the tool or in trying to specify the model correctly, somebody else had already figured it out or written a paper on it. for example, i will definitely be citing this really fabulous journal article on generalized linear mixed models (Bolker et. al, 2009) with not one, but TWO online supplements that are equally fabulous. who knows what i would have done without having all this information at my fingertips!!

part 1 data collection complete

emilee — Sun, 21 Dec 2008 04:04:07 +0000

it’s official! i paid my 100th subject this evening. there were days when i seriously didn’t think i would reach my goal of 100 subjects in part 1 of the experiment. thanks to all the volunteers!! and, on to part 2 (in january). now, if i could just finish this grading…

good news! and other stuff

emilee — Sat, 25 Aug 2007 21:51:18 +0000

I found out yesterday that I’ll be going to Florida in November for the GROUP ’07 doctoral consortium. It isn’t as exotic a location as Rio de Janeiro, but the focus of the conference is more relevant. I was supposed to go to Rio in a couple of weeks for the INTERACT ’07 doctoral consortium, but my passport renewal took an insanely long time (as happened to a lot of people this summer). And, I couldn’t get a visa without a passport. I finally received my passport about a week ago, too late to make the travel arrangements. But, at least now I have it.

Data analysis on the Taboo study has been progressing, slowly. I’ve been meaning to post about it, but every time I think I’ve done enough to sit down and write about, I think of something else to try. The latest ‘something else’ is transcribing one of the other media conditions, both because it will make a more interesting submission to CHI 2008 (in Florence!!), and because I think the results I’m getting suggest that something is going on, but I need more data. I’m just having trouble interpreting the results I have so far, and I’m hoping the additional data will help clear things up. If not, well, I’m really happy to finally have had the chance to analyze this data, and I’ve learned a lot about MANOVA in the past few weeks!

Transcribing the ‘audio-only’ condition has provided me with a few laughs. Here’s an example of one pair who did not previously know each other. The clue-giver (B) clearly has an incorrect model of his partner’s (A) knowledge where Russian composers are concerned. [ listen: AIF 276k ]

B:tchaikovsky wrote a lot of
A:books, novels?
B:tchaikovsky.
A:russian?
B:he’s a musician. musician.
A:oh
B:he composed. it’s like in a theater, but it’s a musical
A:opera
B:not an opera, tchaikovsky was not into opera, but
A:ballet

What you can’t tell from the transcript (without listening to the audio) is that B has a recognizably Russian accent, and seems incredulous that A has never heard of Tchaikovsky. Ahhh, Taboo.

This same pair, however, got one of the most difficult words in just one clue. [ listen: AIF 218k ]

B:it’s an animal. ugly, big red nose and big red butt
A:baboon

Amazing! And, it is things like this that are making the data hard to analyze. So much variability!