Class20.html

Class 20: Probability Demos

Journal Assignment

What are some of the problems of using capture-recapture to estimate census undercount? When possible, make analogies to the problems of using capture-recapture to estimate fish populations, such as birth and death, tags falling off, trap-friendly and trap-shy fish, some species being easier to catch, etc. Are there any problems with the census example that don't have analogies to the fish example, or vice versa?

Remember also that we would like you to include a brief description of your ideas for a final project in your next journal.

Comments on Last Batch of Journals

Joseph read most parts of the journals this time, but Laurie read the Alzheimer's part.

The SAT article, SAT data, chance news article, and cookies data were each given a maximum of 5 points. The Alzheimer's article was worth 4 points, and extra comments were given up to 4 points. If you suspect any part of your journal was not read, please let us know.

Joseph wanted to mention from reading the journals:

weather link

2) someone asked: "how safe ae condoms?" He had an argument with a born-again Christian, and now he's worried.

3) A few people were confusing correlation coeffients with probability. A correlation of 0.5 does not mean that the pattern was just as likely to appear by chance: 0.5 is actually a good correlation. Instead, you would expect a correlation of 0 if there was no relation between the variables and just chance was at work.