Unformatted text preview: 2. Problem 8.9 in Agresti. 3. Problem 9.16 in Agresti. 4. Problem 9.21 in Agresti. 5. Problem 13.8 (parts a–c only) in Agresti. 6. Refer to problems 13.14–13.16 in Agresti. Using width and qualitative color as predictors, ﬁt both a Poisson GLM, and a negative binomial GLM to these data (number of horseshoe-crab satellites), checking for interaction and interpreting the ﬁnal model. Here you should identify which of these models is more appropriate and summarize all the relevant statistical results that support your choice of model for these data. In addition, comment on the goodness of ﬁt of your chosen model and summarize whatever inferences regarding the eﬀects of width and color you feel are appropriate for these data....
View Full Document
- Fall '11
- Statistics, Agresti, Advanced Statistical Applications, negative binomial GLM, relevant statistical results