Psych 7 Week 3
Popular in Intro to Experimental Psych
Popular in Department
This 3 page Class Notes was uploaded by Debbie Chen on Thursday October 23, 2014. The Class Notes belongs to 7 at University of California Santa Barbara taught by PROTZKO in Fall2014. Since its upload, it has received 45 views.
Reviews for Psych 7 Week 3
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 10/23/14
Psych gz Class Notes Introduction to Experimental Psych Week 3 Causality 0 1 rule correlation does not equal causation o Guiliani o How to recognize correlations immediately I People who do Xdon t do X are Y I People who drink wine are happier I People who meditate live longer I People who are taxed more work less I Strength of Association 0 Stronger is better doesn t mean weak is bad can still be causal I Consistency 0 Can be found by other people 0 Replicated I Speci city 0 Narrow down to as few criteria as possible I Temporality 0 Causes come before effects I Doseresponse 0 Fist ghts 8 aggressive 0 People who get in more st ghts should become more aggressive I Plausibility 0 Still a method but too reliant on what you already know I Coherence I Experimental 0 Stomach ulcers scientist tested by drinking bacteria 0 New Criteria I How do we know when something causes a change in something else Counterfactual 0 The most important concept in this class is that of the counterfactual 0 Most causal inference statisticians de ne causal effects as comparisons between what would happen in two or more different states one of which will be factual the others counterfactual 0 So here s the question you ask What would have happened if they didn t get the treatment 0 Example 0 Hypothesis learning to play a musical instrument raises your intelligence I Falsi able 0 Can t compare people who have learned to play a musical instrument to those who haven t I Confounding variables Wealthier 2parent homes I Direction of causality people who choose to play instruments are already smarter to begin with 0 Train people to play a musical instrument and see if their intelligence goes up I Maybe intelligence was already going up Train some to play a musical instrument and don t train others What I want is the counterfactual what would have happened to them if they hadn t been trained 0 What the Counterfactual Does 0 People get treatment then we measure so temporality is still good 0 I m comparing people against their counterfactual selves so I control for all covariates The only difference the independent variable Problem if I train some people to play a musical instrument how can I observe what would have happened if they hadn t been trained 0 Random assignment 0 On average one group is the same as the other 0 One stands as is not the counterfactual for the other 0 Is there anything else 0 How do you de ne your counterfactual group 0 Example does smoking cause lung cancer 0 Must explicitly state the counterfactual o 3 smoking groups I Smoking one pack a day prevents lung cancer I Smoking causes lung cancer compared to not smoking 0 Counterfactual in Action I want to know the effects of meditation on happiness Experimental Group People who signed up for meditation intervention Control Group Other people I nd Does this count as a counterfactual No OOO I Some are in the experimental group other can you come back in 6 months are the control group Wrapup 0 Cannot infer causality from correlational design 0 Confounding variables create an alternate hypothesis for your data 0 It is normally impossible to control for every CV 0 By randomly assigning people to groups you can estimate the counterfactual AND happen to control for all of your CVs