Next week we’ll have Dr. Van Valin as a guest host! We’ll meet at 3:30 in 603 Baldy.
Our topic will be: Exploring Mechanical Turk for a rating study on core co-subordination.
November 12th we will be Skyping with John Olstad at 3:30. Check back here for location.
Anastasia’s presentation on Posture verbs in Russian has been moved to November 19th.
We’ll be in 603 Baldy again, meeting at 2:00.
We got a little carried away with our discussion of Wolff et al last week, and we never got to GLMM part II — so this week is GLMM part II take II. We’ll take an in-depth look at the types of data students are collecting, and try to identify where you can get numbers for analysis.
If you have data you’d like to be considered, feel free to email Kate or Juergen. We also have a discussion board set up on UBLearns (Blackboard) where you can ask stats-related questions.
You may want to look at R before lab, or peruse some discussion on the software at Stackexchange.
This week we’ll use about 30 minutes to finish a discussion on GLMM we started last week. We’ll have brief discussions on what kind of data you can get in different situations. We’ll consider tasks run by various lab members. If you have a task in mind and would like to talk briefly about what kind of data you can get, bring it up!
After that, Sang-Hee Park will be leading a discussion based on “Causers in English, Korean, and Chinese and the individuation of events” by Phillip Wolff, Ga-hyun Jeon, and Yu Li (Lanugage and Cognition, 2009). The paper is available on UBLearns. If you would like a copy and are not part of the UBLearns system, feel free to email Randi Moore.
This week, Juergen and Kate will lead lab. This meeting will focus on using Generalized Linear Mixed-effects Models (GLMM). GLMM model the relationship between a dependent variable and the explanatory variables. We will discuss how to implement these models using a variety of data sets from the lab.
There are two readings for this lab: Johnson (2008) chapters 5 and 7.
If you need access to Blackboard and / or the readings, email Randi Moore.
Hello ST people —
The meeting for this week will be held at 3:30 pm, instead of 2:00, as previously scheduled.
Upcoming Events has been modified to reflect this. All else remains the same.
This week will be devoted to some issues relating to video data.
Issues will include: What kind of metrics do you want to get from a video recording of a task? How do we get those numbers? What in the session is useful, and what is not? How much detail should you strive for? How accurate can you be?
For the majority of the discussion, we will be working with data from Holly (Williams) Keily’s QP pilot study on co-speech gesture accompanying event descriptions. There will also be a brief overview of the project.