Question

2 Which one of the following statements is true about the second step in the process for regression

forecasting?

An analyst must determine whether the relationship is statistically significant based on a t-test.
An analyst must evaluate the model for serial correlation.
An analyst must evaluate the explanatory power of the model.
An analyst must evaluate whether the model is logical.

3 A negative serial correlation exists when a _______________ error is followed by a _______________ error.

4 Which one of the following correctly explains the difference between a trend model and a causal model?

A trend model uses a form of smoothing analysis to project the past time trend forward while the causal model looks at a change in an independent variable that causes a change in a dependent variable.
A trend model looks at the past time trend to apply regression analysis while the causal model looks at a change in a dependent variable that causes a change in an independent variable.
A trend model identifies the factors causing change and places them into a bivariate regression model while the causal model matches the slope of the trend through an independent variable tied to a dependent variable’s change.
A trend model tracks the past time trend and projects it forward while the causal model looks at a change in an independent variable that causes a change in a dependent variable.

5 Why is the first step in the regression model evaluation so important?

We desire the explanatory power of the model to be at least 84% of the variation in the dependent variable.
We want the relationship to be statistically significant at the desired level of confidence.
We would never want to use a relationship that does not conform to business/economic logic.
We need to determine if the Durbin-Watson test is within our range of zero to four to rule out serial correlation.

6 Visualization of data allows you to ____________________.

be as transparent to management as required
see stark differences that would not be apparent from the descriptive statistics
better understand if you need more data
more clearly identify the dependent and independent variables

7 What is heteroscedasticity?

When the error terms in the population regression have a constant variance across all values of the independent variable.
When the scatter plot of residuals falls in a horizontal band.
When the standard errors of the regression coefficients may be underestimated causing the calculated t-ratios to be larger than they should be.
When the scatter plot of residuals falls in a vertical band.

8 What assumption does the causal model make?

Changes only occur in the variable to be forecast, but that change is not related to the independent variable.
No changes occur.
Changes in the independent variable will cause a change in the variable to be forecast.
Changes in the dependent variable will cause changes to other dependent variables.

9 Heteroscedasticity is more common with _______________ data than with _______________ data?

10 What is the primary purpose of the third step when you are evaluating a linear regression model?

the process for regression forecasting