Papierkorb

Question 1

Pearson’s Correlation Coefficient

Answer 1

Measures the association between two continuous variables

correlation does not equal causation

Pearson’s r E [-1,1]
is positive when dv & iv increase together
is negative when dv & iv decrease together

=1 when variables perfectly + correlated
= 0 when no correlation
= -1 when variables perfectly - correlated

Question 2

Predicting values

Exercise: You want to predict the gini index for a voter turnout of 40 percent.

Intercept: 56.1
coefficient of turnout: -0.24

How do you calculate that?

Answer 2

1. yi = intercept (a) + coefficient (b) * xi + error (the regression equation)

-> plug in the variables and coefficients

gini i = 56.10 - 0.24 * vote turnout + error

2. plug in the value of your variable for which you want to predict (-> you can leave out the error term here)

gini (hat) = 56.1 - 0.24 * 40
= 46.5