Question

e Pas 2. If SSE is near zero in a regression, conclusion will be that the proposed model probably has a. too poor a fit to be useful b. near perfect fit c. a very strong positive linear relation d. a very strong negative linear relation 3. A residual/error shows the difference between Yactual and Yestimated on a standard scale. True False 4. In a simple regression, the F statistic is calculated by taking the ratio of MSR to the MSE. True False 5. A negative value for the correlation coefficient () implies a negative value for the slope (bi). True False 6. There are usually several possible regression lines that will minimize the sum of squared errors. True False 7. The variable used to predict another variable is called the A. response variable. B. regression variable. C. independent variable D. dependent variable.

Answer 1

2) option b) near perfect is correct.

SSE =Σ(y; – 9:02

Thus if our model is best fit to the data then SSE will be close to 0 i.e. minimum because $\hat{y_i{}}$ will be close to y_i

3) True

because residual is defined as

(Yactual - Ypredicted)

4) True

F = MSR/MSE follows F distribution with (degree of freedom due to regression, degree of freedom due to error)

MSR = SSR/degree of freedom due to regression

MSE = SSE/degree of freedom due to error

5) False

negative correlation means increasing 1 unit in input variable will show decrease in output variable by correlation r. if there is negative correlation it doesn't imply slope to be negative.

because simple linear regression model is defined as

y = a+b*x+error

6) False

In linear regression we estimate the unknown parameters using the principal of least square which gives the best estimates of unknown parameters which will result in minimum error. Therefore there is only one possible line which minimize the error.

7) option c) Independent variable is correct