Starting running my user study (the research equivalent of a how-to project? see Research section). Apparently it's interview week at the ArticuLab, so life is a little frenzied for everyone. It's also Max's and Linh's (fellow DREU interns) last week in Pittsburgh, which is crazy--feels like we all just arrived a couple days ago.
Things I Learned
- animal crackers are hard to decipher
- how to rate rapport
- in Two Truths and a Lie, the first statement tends to be the lie (at least with ArticuLab interns)
- people don't seem to notice a 1-2 second difference in response time
- homemade waffle cones are a big thing
Officially started conducting my user study on Tuesday. The day before, I did a runthrough with Florian, during which he caught an inordinate amount of holes in my execution (i.e. potential questions that I didn't answer and specifications I didn't address. Can the agent recommend any existing movie? Can you request multiple actors at once? (answers: only ones released within the last five years, and no)).Florian also gave me tons of helpful suggestions on how to improve/simplify/organize my explanations. In other words, I now owe him 2834901 times over. I'm also now wondering if I flunked all those how-to projects in grade school* or if I'm just mortifyingly out of practice?
At any rate, it was cool to get the hang of (hopefully) properly running an experiment--and it felt super professional to walk people through consent forms, explain procedures, and slide the conference's sliding door shut so they could talk to the agent privately (even if their voices were audible from outside). A lot of things, like differential equations and parallel parking and Chinese leapfrog, only really click once you actually do them for yourself, and designing/executing a user study is definitely one of those things.
In other news, I helped annotate rapport for Jingya's project, which was a welcome change after, like, my third iteration of the user study. (Figuring out how to explain everything was interesting, but parroting that explanation, not so much.)
* seriously, though, those were awful. I remember doing one on making pizza and literally having to bring in measuring cups and bowls and whatnot to demonstrate to the class, sans ingredients, in real time. Almost as bad as the time I lost my chess pawns to a social studies diorama.
Samantha went over what the ArticuLab looks for in our final presentations; we also chatted about some other helpful hints on showcasing your work upon completion.
Linh and Max (fellow DREU interns) gave their final presentations.
These are more or less lifted from Bitbucket, where I keep a README of these things.
User study planning/changes, runthrough.
User study prep; ran user study; Davos rapport annotation.
User study prep; ran user study; Davos rapport annotation; website.
Ran user study; rdg group.
Mtg with Lauren; lab mtg; mtg with Timo; ran user study; organized data.
As a send-off for Max and Linh, us interns + Zhen walked downtown? easttown? to get (fantastic) poke bowls for dinner and homemade ice cream/cones for dessert. Afterwards, we headed back to the lab, where Linh and Jingya worked on calculating utopy, the rest of us watched a Black Mirror episode on the ArticuLab's fancy new Mirror surface, and everyone doodled on said surface.
- finally made it to the Andy Warhol Museum, which was great (also, I adore museums in general)
- went kayaking (in the Allegheny River) for the first time & learned that I have zero upper body strength
- checked out the Monongahela Incline
- also, hibernated for a good portion of Saturday