SparkNotes

For those of you who missed classes this week, here's the rundown:

American Politics
Polling is great, but you have to be careful about survey questions. Not only can there be statistical errors in extrapolating sample data to a larger group, but there are other errors to look out for: are your respondents really random, or are you isolating some people by the way you find your participants? Are you suggesting that a certain response to the question is preferred?

Even if not, some people lie, perhaps to sound better than they are. Always give people a way out. Turns out, the survey we filled out the class prior had some experiments in it. In one experiment, we rated political groups, some of which didn't exist. About 2/3 of the class had strong opinions of fictional groups; that's an invented opinion, a lie.

Another question asked us about our approval of Massachusetts criminal procedures. Half of us were given the question with the additional sentence: "Write 'Don't Know' if you do not know enough to have an opinion". Those presented with the way out almost all took it; those who weren't explicit told to write that ended up inventing drastic approval ratings. Interesting.

Spanish
Commas serve as decimals for Spanish speakers. 4.6 becomes 4,6.

LOST
Television can be analyzed as literate, much as a Shakespearean play can. Not all shows, of course, but ones with a storyline, symbolism, and meaning. Unlike most shows, LOST has a big storyline: you can't just watch a random episode here or there. The character development, recurring motifs, and philosophical messages will be explored in our course. Homework: Watch 2 specifically chosen (and thematically related) episodes on Hulu. Then read specific commentaries. Then write a short paper. This class rules. We even break for food halfway through.