Reference no: EM133939044
Questions:
1. Describe what is meant by "model," as it is used in epidemiology and biostatistics.
2. What is being "predicted" in a logistic regression (that is, what form is used for the outcome or dependent variable).
3. What is being "predicted" in a proportionate hazards survival model (that is, what form is used for the outcome or dependent variable)?
Read through the provided paper: Sir Richard Doll: Cohort studies: history of the method. 1. Prospective cohort studies.
4. In figure 4 he uses a somewhat different definition of the term "cohort" from theone we've been using in study design. In this example, he is comparing the mortality experience from lung cancer of men born during different "eras" or "periods": 1945, 1940-44, 1936-39, 1931-35, 1021-30, 1911-20. Each of these groups may be thought of as a cohort.
The cancer experience differs by age and by period of birth for each of these cohorts. Describe the differences observed in groups (cohorts) A through F.
5. Many of these studies that Doll describes were conducted before the computer era. What are the major improvements that computers offer in the analysis of studies of this sort (or really, studies in general)?