Week 13

Summary

Penultimate week at the ArticuLab: did user study things, glared at a lot of statistical terms, and tried to pull my internship work (i.e. analysis tool and dataset in progress) together and start tying loose ends.

Things I Learned

  • The Porch has half-price pizza on weekday nights
  • at least ten different types of statistical tests
  • university bookstores' greeting cards sections are comprised of half birthdays, half consolations, and no thank yous whatsoever

Research

(Aug 14 update on last week's mention of user study repetitiveness: If I have to run the experiment one more time, I actually might explode.)

If I had to break down my week into three tasks, I'd say i) continued running my user study, ii) spoke with Florian about different statistical tests and which one would fit my dataset best (more on all that below), and iii) adjusted my analysis tool so it would process my data properly.

Since Timo's going to Interspeech (and thus won't be here for my last week), we also talked "offboarding" and the logistics of working remotely.

Statistics & Me: Life and Love with the World's Worst* Subject**

#2 in a nutshell: basically, I spent a whole lot of time puzzling over statistical tests and wishing I could better understand them. I'm embarrassed to say that the only thing I remember from high school stats is that time I stayed up all night using technology (to analyze data on people staying up all night to use technology) and ate an entire bowl of pudding.

By "an entire bowl," I mean that I did indeed make a mixing bowl's worth of pudding at twelve in the morning and proceeded to empty that entire vat with a spoon and a spatula.

I also realized that I haven't touched my graphing calculator since high school stats, during which the most significant thing I used said calculator for was entering in the measurements for one hundred different blades of grass (that's right, not even actual graphing). I have an unhealthy mistrust of technology, specifically calculators***, which is probably not great for a computer science major and for anyone analyzing data in general.

Anyway, this is (kind of?) terrible, but I ultimately gave up on the stats proofs (sorry, eigenvalues) and decided to be content with knowing which test to use when and what the outcome of each test means. I do eventually want to fully understand how smart people came up with all these means of analysis, but for now, I suppose doing the actual analysis is more important.

* but most applicable

** not really, but still

*** all my kitchen appliances have names, yet my graphing calculator remains unpersonified; this speaks volumes about my priorities in life


Meetings

Lab Meeting

Sarah (leaving Friday) and Jingya (leaving next Tuesday) delivered their final presentations.

InMind Meeting

Timo called a meeting with Yoichi, Florian, Sam (new masters student), and me to talk about prepping for some demos over the next week slash upcoming deadlines over the next month.

Daily Logs

These are more or less lifted from Bitbucket, where I keep a README of these things.

Mon 8/14

InMind mtg; website; ran user study; organized data.

Tues 8/15

Looked at data; updated analysis tool; SARA face training session.

Wed 8/16

Mtg with Florian; looked at data; fixed analysis tool; ran user study.

Thurs 8/17

Altered analysis tool; cleaned up data; stats; ran user study.

Fri 8/18

Lab mtg; cleaned up data; stats; InMind mtg.

Not Work

SARA (Team) Dinner

To celebrate the end of Jingya's internship (she leaves next Tuesday) and also just eat, our team went to shabu-shabu (hot pot) ((which we've frequently discussed during our daily lunches)) for dinner on Friday! Thumbs up for food and team bonding.

Miscellaneous

  • went school supply shopping (check out that sibilance) at Fancy Target
  • I'm sure I read, but I don't remember what
  • cooked food that bears slightly more resemblance to actual meals than usual

Template by Codrops | Last updated 02 Sept 2017