Question 1

p-values decrease as the error variance decreases for the same sample size and effect size. Select one:

True

False

Question 2

Explaining and prediction have the same objective. Select one:

True

False

Question 3

Explaining relates to testing a hypothesis that a factor Xi is associated with the outcome variable Y.

Select one:

True

False

Question 4

Maximizing R Square to find a model can lead to over-fitting.

True

False

Question 5

t-tests in regression require that the error is normally distributed.

True

False

Question 6

Parsimony principle says that one should strive for the fewest number of predictors that provide about the same prediction accuracy as more predictors would. True

False

Question 7

When building a model for prediction we should split the data into training and validation to guard against over fitting. Select one:

True

False

Question 8

Which is not a metrics used when building regression models with many predictors?

Select one:

A.

R Square

B.

Adjusted R Square

C.

Akaike Information Criterion (AIC)

D.

Bayesian Information Criterion (BIC)

E.

Mallows's Cp

Question 9

Which condition is not required for building a linear regression model to explain the effect of factors?

a.

Normality of residual errors.

b.

Homoscedasticity: The variance of residual is the sameforany value of X.

c.

The relationship between X and the mean of Y islinear.

d.

Independence: Observations are independent of each other.

e.

All are conditions.