# Linear regression, multiple regression, time series

Please see attached file

First some questions about simple, linear, bivariate regression.

1 Of these correlation coefficient values, which represents the least strength of correlation?

A +0.4

B +0.1

C -0.2

D -0.5

2 The t test for significance of the correlation coefficient has how many degrees of freedom?

A n

B n-1

C n-2

D n-3

3 Which is NOT a good indicator of the goodness of fit of a regression equation?

A Correlation coefficient

B Coefficient of determination

C F statistic

D Equation slope

4 Which of these is NOT computed from the regression's standard error value?

A Confidence interval for slope

B Confidence interval for intercept

C Residuals

D Prediction interval for Y

5 Which is NOT an assumption made about the random errors in a regression model?

A The errors are normally distributed

B The errors have constant variance

C The errors are independent

D The errors are small

Now for multiple regression.

6 Suppose these models all have about the same R-squared value. Which is best?

A Bivariate regression with one independent variable

B Multivariate regression with two independent variables

C Multivariate regression with three independent variables

D Multivariate regression with four independent variables

7 The F test for significance of a multiple regression tests which null hypothesis?

A R2 = 0

B rho = 0

C beta 1 = beta 2 = beta 3 = ... = beta k = 0

D e1 = e 2 = e 3 = ... = e n = 0

8 Heteroscedasticity has to do with...

A Whether variance is constant

B Whether variance is autocorrelated

C Whether variance is normal

D Whether variance is known

And a couple of questions about time series.

9 Which is NOT a measure of fit for a time series trend model?

A MAD

B MUD

C MSD

D MAPE

10 Which is NOT a moving average trend analysis method?

A TMA

B CMA

C Exponential smoothing

D Deseasonalization

#### Solution Preview

Please see attached file

For each question, place your cursor in the yellow highlighted cell and click the left mouse button. Then choose the correct answer letter from the drop-down list. Note: This is intended to be a learning exercise. You are welcome to refer to your text, research these questions on the web, or discuss them with other class members. But turn in your own answer spreadsheet -- the one with your name shown above.

First some questions about simple, linear, bivariate regression.

1 Of these correlation coefficient values, which represents the least strength of correlation? B

A +0.4

B +0.1

C -0.2

D -0.5

Answer: +0.1

Strength of correlation is indicated by coefficient of determination; coefficient of determination=correlation coefficient ^2. The least value is for r=0.1

r= r^2=

0.4 0.16

0.1 0.01

-0.2 0.04

-0.5 0.25

2 The t test for significance of the correlation coefficient has how many degrees of freedom? C

A n

B n-1

C n-2

D n-3

Answer: n-2

3 Which is NOT a good indicator of the goodness of fit of a regression equation? D

A Correlation coefficient

B Coefficient of determination

C F statistic

D ...

#### Solution Summary

Answers Multiple Choice Questions on simple, linear, bivariate regression, multiple regression, time series.