Class 2: Correlation in not Causation Flashcards
Correlation
Is a statistical technique which tells us how strongly the pair of variables are linearly related and change together.
- shows relationships exists
- is something which we think, when can’t see the hidden factor
Causation
Any change in the value of one variable will cause change in the value of another variable. Which means one variable makes other to happen.
-referred as cause and effect
Counterfactual
Are conditions that would have been true under different circumstances
- A hypothetical situation that would’ve been true in similar but different circumstance
Rothmans Definition
“ A cause of a Specific disease event[is] an antecedent event, conditions or characteristics that was necessary for the disease at the moment it occurred given the other conditions are fixed.”
-best way to have counter factual in humans
Randomized control Trial (RCT)
-Lots of people
-treatment random
-double-blind
Can’t have CRT
-treatment unethical
When impossible to have placebo or control group
Rathfords Hills Causal Consideration
- 1965
- used by World Health Orgoniaztion
1) strength
Larger associations are more likely causal
2) consistency
Same conclusion across studies, samples, population places
-same results
3) Coherence
Coherence between epidemiological/human and experimental findings( a special case of consistency)
-populations, places,species
4) Specificity
No other explanation given consideration of alternative
- ex: why that disease occurred
- many variables happening( in life)
5) Temporality
Effect( health outcomes) occur after cause(exposure)
- at a later time
6) Biological Gradient
Dose- response relationships, greater exposure, greater effect
7) Plausibility
Biological mechanisms possible and well evidenced
-DNA changes as mutations
8) Experiment/Reversibility
Demonstrating that outcome happens in preserve but not absence of exposure/outcome reduced when exposure removed
Pérceles
The dose make the poison
