PLCY 460 NEED TO KNOW TERMS

Moderator Relationship

A variable that directly influences the relationship between an independent and dependent variable.

(The relationship between education and income is affected by the variable of gender)

Confounder Relationship

A third variable that influences both independent and dependent variables. This third variable might distort the relationship between the independent and dependent variables.

(Education affects income, but the variable of family wealth has an effect on BOTH education and income seperately.)

Mediator Relationship

There is a better, third variable that explains how the independent variable interacts with the dependent variable.

The independent variable effects BOTH the third variable and the dependent variable.

(Education impacts Income, but education also impacts Job Skills. Job skills also has an impact on income).

Unrelated relationship

A 2nd variable impacts the dependent variable, with no relationship to the independent variable.

Education impacts income, but an ebay addiction has no relation to education but still effects overall income).

What is the first step in establishing causality?

Temporal Precedence.
Establish that this being a causality makes sense. An event in the present cannot affect data from the past.

What is the second step in establishing causality?

Ensuring that the data backs up a potential causality. This means that there is conclusive evidence in the data that there is a relationship between the two variables.

What is the third step in establishing causality?

No confounding variables. Ensure that there are no other variables present that could explain why these two variables have this type of relationship.

What is endogeneity?

When a variable is correlated with something in the error term. Because it is INFLUENCED by things that the model cannot account for, the existence of endogeneity CANNOT establish causality.

Definition of ordinary Least Squares

The minimized sum of squared residuals from the data point to the estimated regression.

What is Standard Deviation

The spread of the actual factual data values

Standard Error

The variablity/uncertainty of the estimates.

What represents true mean of population?

Mu (letter U with a tail)

What represents true standard deviation of a population?

Sigma (O with the line on top)

What represents the SAMPLE MEAN

X bar

What represents the SAMPLE standard deviation

S

What does R-Squared mean

Proporiton of the variation of Y around its mean explained by the model (X values)

Used to determine the accuracy of the model

Root MSE (sigma hat)

How far the points are, on average, from the line

Used to determine the accuracy of the model