Term
|
Definition
1)Model is Linear in the Parameters
2) The X Variable is Uncorrelated with the U disturbance term
3) E(u/Xi) = 0
4) Each ui variance is homoskedastic (constant)
5) No correlation between two error terms
6) No specification bias (all relevant variables included)
7) ui is normally distributed (mean 0, variance sigma^2) |
|
|
Term
|
Definition
| min(b1,b2,b3)Sigma(e^2) = RSS = Sigma(Yi - b1 - b2x2 - b3x3...)^2 |
|
|
Term
| Goodness of Fit/R^2 equation |
|
Definition
|
|
Term
|
Definition
|
|
Term
| Partial Regression Coefficient |
|
Definition
| Reflects the (partial) effect of one explanatory variable on the mean value of the dependent variable when the values of other explanatory variables included in the model are held constant. |
|
|
Term
| Coefficient of Determination |
|
Definition
| The proportion in variation of the dependent variable explained by all independent variables in the regression model. |
|
|
Term
|
Definition
| Multiple (exact) linear relationships exist between variables in an estimated multiple regression function. |
|
|
Term
|
Definition
|
|
Term
|
Definition
| lnY = B1 + B2LnX (elasticity) |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| Y = B1 + B2X + B3X^2 + B4X^3 |
|
|
Term
| Dubin-Watson Statistic EQN |
|
Definition
d= (Sum(Et - E(t-1))/(Sum Et^2) or
d approx = 2(1-p) (p: Covariance of Et & Et-1 |
|
|
Term
| Correct Variance Methods (3) |
|
Definition
1) Divide through by SQRT variance
2) Divide through by SQRT of X
3) Divide through by X |
|
|
Term
| Dummy Variable Regression Iterations (4) |
|
Definition
1) Coincidental Regression
2) Concurrent Regression
3) Parallel Regressions
4) Dissimilar Regressions |
|
|
