Correlation
Correlation
In empirical data, when A is higher, B is systematically higher (or lower) --- RANK ORDER ! Attributes of one variable are associated with attributes of another variable
Rank Order: putting the highest number of drink participants in order so it would go participant 3,7,10,,2,9,5,6,,8,4,1 Among the same people the amount of time spent in the gym is participants 3,7,10,9,2,5,6,8,1,4
An empirical relationship between two variables such that
# of drinks
# of Hours spent in Gym
Causation
Causation
A causes B If we can show increase( B will increase An impact of A on B E.g. Class size & Test score
or decrease) A
,
1
Correlation vs. Causation
CORRELATION IS NOT CAUSATION
From correlation to implied causation… to policy proposal
Evidence
Correlation from survey of teenagers Implied causation If parents would have dinner with their children more often, the children would use drugs less “The most effective thing we can do” Family dinners cause (in part) the extent of teen drug abuse Family Day—A Day to Eat Dinner with Your Children
Conclusion
Negative
Family dinner
Drug Use
Policy proposal
In conclusion: to keep the teenagers from using drugs we will be having more family dinners
Causation concluded
Figure 10.1 (top) p. 332
Causation implied:
If we persuade parents to have more dinners with their