SOSC 2500 Lecture Notes - Lecture 12: Null Hypothesis, Test Statistic, Bulgarian Lev
Document Summary
An association (or correlation) between two variables tells us nothing about why the variables are associated or if one causes the other. Causal inferences (inferring causation) is only possible in highly controlled experi- mental research situations. When we expect two variables to be correlated, and we find that they are, our brain often automatically makes causal inferences, be aware of this, and try not to do it! Causation is something that largely exists in our minds, it is a model or an explanation that we sometimes mistakenly impose on our data or results. Your brain is very good at inferring causation, it is so good at it, that it infers causation where none exists, and it creates patterns that do not exist. Scatter plots allow you to simultaneously view the values of two variables on a case- by-case basis. It is a visual representation of the association (correlation) btwn two variables. Associations can be linear, curvilinear, or non-existent.