Week Nine

Question 0

What does the OLS model do?

Answer 0

It can calculate what the model will predict for each response or case.

Question 1

What model builds a description of the world?

Answer 1

The linear regression model.

Question 2

In the OLS model what does it mean if the result is:

A. Positive?
B. Negative?

Answer 2

A. The observed value is greater than the predicted value.

B. The observed value is smaller than the predicted value.

Question 3

What are the five requirements of the OLS model?

Answer 3

1. Exogeneity of the explanatory variable.
2. Linearity and additivity of the relationship between dependent variables and explanatory variables.
3. Statistical independence of the errors.
4. Homoscedasicity
5. Normality of the error distribution.

Question 4

What are the five issues that can occur with residuals?

Answer 4

1. Outliers bias in estimates of coefficients
2. Non-normality of residuals
3. Heteroscedasticity
4. Omitted variable bias
5. Multicollinearity

Question 5

How can we tell if there is multicollinearity in the variables?

Answer 5

When the r-squared is high but the t-statistic is low.

Question 6

Why do we standardise variables?

Answer 6

To overcome issues with comparing variables that are measured in different units. The variable becomes expressed in units of standard deviation.

Question 7

How do we standardise?

Answer 7

Subtract the mean from each value and then divide by the SD.

Question 8

What is centring?

Answer 8

An alternative to standardising. We use centring when we want to keep the variables in its original units, but the mean is set to 0.

Question 9

How do we centre?

Answer 9

Subtract the mean from each value.