This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: STAT 8630 Final Exam Due Wednesday, May 6 SHOW ALL WORK Name: Do not discuss this exam with anyone other than your instructor. You may use any notes, books, journal articles etc., but you are not to collaborate! 1. Last semester in STAT 8620 we considered an example involving counts of bicycle traffic. Recall that the data came from Gelman et al. (1995), who reported the results of a survey of bicycle and other traffic in the neighborhood of the UC- Berkeley campus conducted in 1993. Sixty streets were selected at random, with a stratification into 3 levels of activity and whether the street had a marked bicycle lane. The counts observed during one hour are shown below. Note that for two streets, the data were lost. Last semester we examined the effect of bike lanes on bicycle traffic and whether that effect differs by type of street (residential, side, main). In this problem we now account for effects of these factors on both bicycle traffic and other vehicle traffic, and account for possible correlation between the two types of traffic. One way to build correlation into a model for two response variables is through a shared random effect model. The selection models we discussed and illustrated on 1 the schizophrenia data were based on this idea. In the present context, consider a model in which, conditional on a street-specific random effect, the bike traffic count and other vehicle traffic count are independent, Poisson random variables...
View Full Document
This note was uploaded on 11/13/2011 for the course STAT 8630 taught by Professor Staff during the Fall '08 term at University of Georgia Athens.
- Fall '08