Gov 2001,
Wow! Almost the entire class responded to the survey expressing interest in
coming to one of the optional sections! That's so cool! You guys are the
best. I wish we had time to do one lecture on all the major vote getters,
but we really don't (and I imagine interest would wane after the third or
fourth "final section").
So there were three topics which each got more than 15% of the vote. I've
put up a new poll (same place) that asks you to pick your favorite of these
three categories. We are doing this for two reasons: (1) We want the topic
people are most interested in, not just the one the most people will attend,
(2) some people mentioned that they didn't know what all the topics meant.
So here is a brief summary of the top 3:
Bayesian Statistics: We will cover the Bayesian approach to data analysis
focusing on the differences with the frequentist statistics covered in
class. We will outline MCMC methods and provide some applications for which
Bayesian methods might be useful. Finally we will direct you to resources
for more information and take questions.
Spatial Statistics: We will cover spatial data and basic visualization in R,
the spatial autoregressive model and how to choose spatial weights (with a
focus on the international relations context but its generally applicable).
We will then direct you to the canonical resources for more information and
take questions.
Applied Text Analysis in R: We will cover two kinds of text analysis on the
last day of lecture. In this section, we will show you how to apply some of
these methods in practice. We will provide a quick summary of the
literature and then mostly show code for exploring textual data, visualizing
textual data, supervised learning (document classification) and unsupervised
learning (clustering).
If you have any questions feel free to shoot me an email. Go vote on your
favorite. I'll announce the winner next week.
Also thanks to everyone who suggested additional topics: anchoring
vignettes, GIS, IV, regression discontinuity, structural equation modelling,
selection models and time series were all suggested. If you want to talk
about any of these things- come see me in office hours. I can direct you to
useful resources.
Brandon
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l