120 likes | 514 Views
Understanding Observational Studies. Lecture 12 Secs. 3.4 Wed, Feb 4, 2004. Types of Observational Studies. Retrospective study – The subjects are grouped by the response variable. Then the explanatory variable is measured for each group and compared.
E N D
Understanding Observational Studies Lecture 12 Secs. 3.4 Wed, Feb 4, 2004
Types of Observational Studies • Retrospective study – The subjects are grouped by the response variable. Then the explanatory variable is measured for each group and compared. • Prospective study – The subjects are grouped by the explanatory variable. Then the response variable is measured for each group and compared.
Example • Suppose researchers want to study the relationship between the amount of fat in one’s diet and the incidence of heart disease.
Example • A retrospective study might begin by dividing the subjects into two groups. • Those currently with heart disease. • Those currently without heart disease. • Then they would investigate the subjects’ past eating habits.
Example • A prospective study might begin by dividing the subjects into two groups. • Those currently eating a low-fat diet. • Those currently eating a high-fat diet. • Then they would observe the subjects over the next several years to see which ones develop heart disease.
Retrospective Studies • In a retrospective study, the response variable’s value is often its present value. • The value of the explanatory variable is determined by looking at past records or through interviews of subjects.
Prospective Studies • In a prospective study, the explanatory variable’s value is often its present value. • The value of the response variable is determined by tracking the subjects over weeks, months, or years into the future.
Let’s Do It! • Let’s do it! 3.4 – Left versus Right. • Major Principle in Statistics: The larger a sample is, the more likely it is to be representative of the population. • How large is the sample of left-handers?
Difficulties in Observational Studies • In an observational study, • The researchers cannot control the explanatory variables. • The researchers cannot control the confounding variables. • In the earlier example, is it plausible to believe that more of those on a high-fat diet would also smoke?
Assignment • Page 142: Exercises 6 – 11. • Page 167: Exercises 30 – 32.