Quantitative Methods Forum ("Quorum") is a weekly discussion group and workshop geared toward all types of linguistic data analysis. We welcome linguists of all stripes, as well as suggestions for specific discussion topics and requests for feedback on all kinds of data. All are welcome to join in the fun, though we do recommend taking Ling c160 (Quantitative Methods) and/or having some basic knowledge of statistical concepts before you arrive.
We are also working on updating and expanding the Lab Wiki. If you are a Berkeley Linguistics grad student and would like to consult or contribute to the Wiki, you can access it here. (First time users may need to contact systems analyst Ronald Sprouse to be added.)
More linear modeling, examining influential cases more in depth.
Some concepts in linear model evaluation. What is influence vs. leverage? What should we watch out for in removing outliers from data? Follow-up on Hay & Bauer (2007) and an R script for data simulation posted to the Lab Wiki.
Correlation. What about Hay & Bauer (2007)'s finding that population size is loosely (though significantly) correlated with phonological inventory size? What happens if we try to simulate their data?