mgt 6203 midterm 1 UPDATED ACTUAL
Questions and CORRECT Answers
Explained Deviation - CORRECT ANSWER - y hat - y bar
Total Deviation - CORRECT ANSWER - y-ybar
Unexplained Deviation - CORRECT ANSWER - y-yhat
SST - CORRECT ANSWER - SSE + SSR. - sum of (yi - y bar)^2
SSE - CORRECT ANSWER - sum of (yi - y hat) ^ 2
SSR - CORRECT ANSWER - sum of (y bar - y hat) ^ 2
R^2 (coefficient of determination) - CORRECT ANSWER - 1 - SSE/SST = SSR/SST
Adjusted R^2 - CORRECT ANSWER - 1 - (SSE/(n-p-1)/(SST/n-1)
R^2 = 0 - CORRECT ANSWER - X values account for no difference in Y values
R ^2 range - CORRECT ANSWER -0-1
H0 for Regression - CORRECT ANSWER - Parameters is 0 - this implies that low
probability for analysis is good
F-statistic - CORRECT ANSWER - Used to determine for R^2 if model is good or bad
(with degrees of freedom) for regression
, Common Problems for Linear Regression - CORRECT ANSWER - 1. Non-linearity
2. Correlation of Error terms
3. Heteroscedascity
4. Outliers
5. High Leverage Points
6. Collinearity
Non-linearity - CORRECT ANSWER - Non linear data set
Correlation of Error terms - CORRECT ANSWER - Error terms impact other error terms
(means there is auto-correlation)
Heteroscedascity - CORRECT ANSWER - Non constant variance
Outlier - CORRECT ANSWER - A value that "lies outside" (is much smaller or larger
than) most of the other values in a set of data.
High Leverage Points - CORRECT ANSWER - Multiple outliers in their own "cluster"
that skew results
Cook's Distance - CORRECT ANSWER - Gets rid of outliers and high leverage points
Influential Point - CORRECT ANSWER - An outlier is this if when removed it greatly
changes the fitted model
Mulitcollinearity - CORRECT ANSWER - Variables are linearly related. Can use
Variance Inflation Factors (VIF) to detect
Questions and CORRECT Answers
Explained Deviation - CORRECT ANSWER - y hat - y bar
Total Deviation - CORRECT ANSWER - y-ybar
Unexplained Deviation - CORRECT ANSWER - y-yhat
SST - CORRECT ANSWER - SSE + SSR. - sum of (yi - y bar)^2
SSE - CORRECT ANSWER - sum of (yi - y hat) ^ 2
SSR - CORRECT ANSWER - sum of (y bar - y hat) ^ 2
R^2 (coefficient of determination) - CORRECT ANSWER - 1 - SSE/SST = SSR/SST
Adjusted R^2 - CORRECT ANSWER - 1 - (SSE/(n-p-1)/(SST/n-1)
R^2 = 0 - CORRECT ANSWER - X values account for no difference in Y values
R ^2 range - CORRECT ANSWER -0-1
H0 for Regression - CORRECT ANSWER - Parameters is 0 - this implies that low
probability for analysis is good
F-statistic - CORRECT ANSWER - Used to determine for R^2 if model is good or bad
(with degrees of freedom) for regression
, Common Problems for Linear Regression - CORRECT ANSWER - 1. Non-linearity
2. Correlation of Error terms
3. Heteroscedascity
4. Outliers
5. High Leverage Points
6. Collinearity
Non-linearity - CORRECT ANSWER - Non linear data set
Correlation of Error terms - CORRECT ANSWER - Error terms impact other error terms
(means there is auto-correlation)
Heteroscedascity - CORRECT ANSWER - Non constant variance
Outlier - CORRECT ANSWER - A value that "lies outside" (is much smaller or larger
than) most of the other values in a set of data.
High Leverage Points - CORRECT ANSWER - Multiple outliers in their own "cluster"
that skew results
Cook's Distance - CORRECT ANSWER - Gets rid of outliers and high leverage points
Influential Point - CORRECT ANSWER - An outlier is this if when removed it greatly
changes the fitted model
Mulitcollinearity - CORRECT ANSWER - Variables are linearly related. Can use
Variance Inflation Factors (VIF) to detect
