{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Section2.6posted

# Section2.6posted - ST OR 155 Section 2 T hur sday Febr uar...

This preview shows pages 1–6. Sign up to view the full content.

STOR 155, Section 2 Thursday, February 4, 2010 Section 2.6

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Section 2.6: The Question of Causation Causation Common response Confounding Establishing causation
Association does not imply causation! Some observed associations (all positive): Religious-service attendance and longevity Years of education and income SAT scores and first-year college GPA Smoking and lung cancer Marital status and income (married men earn more) How much of each of these is explained by direct causation , and what else could explain them?

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Explaining associations Example 2.38 : x = mother’s BMI (body mass index), y = daughter’s BMI. There is a positive association: r = +.507 , r 2 = 25.7% . Body mass is hereditary; daughters get half their genes from their mothers. So there is at least some direct causation . But this explains at most 25.7% of the variability in daughters’ BMI. The rest is due to other factors. Some possibilities for these other factors: We need to think about the how these other factors might affect daughters’ BMI, in order to understand the association more fully, and also in order to design further studies if necessary.
A simple explanation (if we can establish it): Direct Causation. Changes in x are causing changes in y. Another simple explanation (if we can establish it): Common response. A lurking variable z is causing changes in both x and y . Explaining associations Dotted lines indicate associations to be explained.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}