Quant Final

Question 1

Assumptions of Linear Regression

Answer 1

Linearity: The relationship between the dependent and independent variables is linear.
Independence: The observations are independent of each other.
Homoscedasticity: The variance of the errors is constant across all levels of the independent variables.
Normality: The errors follow a normal distribution.
No multicollinearity: The independent variables are not highly correlated with each other.
No endogeneity: There is no relationship between the errors and the independent variables.

Question 2

effects of non-independent observation when using ordinary regression?

Answer 2

More type I errors.

if kids were randomly assigned to schools and there wasn’t any school or location effect, you’d see distributions around a similar mean

Question 3

examples of non-independent observations

Answer 3

students in a classroom

patieints w/ different clinical psychologists

repeated reaction time trials

assessments of same people across time

dyads in a relationship

members of a family

Question 4

Fixed Effect

Answer 4

predictors with well defined, irreplaceable categories (treatment v. control)

specific levels are of interest

single estimate in model

Question 5

Random effect

Answer 5

group will become predictor

levels are now replaceble.

Levels are random sample from a large population

estimates vary for different clusters/group, predictor as representative of larger population.

Question 6

Components of Generalized Linear Model

Answer 6

Systemic: RHS beta0 + beta1*x1

Random: distribution (normal, binomial, Poisson, Gama). answer to Y ~ ?. It’s a declaration of how data are distributed

link function: function applied to systematic component so that it’s applicable/comparable to the data.