Regression+Assumptions

Regression Assumptions (Regression Analysis | Multivariate Methods) (GWU EMSE-271)
Index | Topics (Logical Lectures) | Lectures | Problems | Readings | Nomenclature | Concepts

For regression analysis (Lattin, Chapter 3):
 * "The matrix X fixed" (not stochastic) "with full rank.
 * "The error (term) "is independently and identically distributed with mean 0 and variance sigma squared." - Lattin

EMSE 271, Fall 2009  "The matrix is of full rank: There is no perfect redundancy in the matrix. Neither of the columns can be written as a linear combination of the others

"The matrix is fixed: They are not stochastic. When is fixed, it cannot be correlated with the error term.

"The error term is independently and identically distributed with mean 0 and variance sigma-squared." - EMSE 271, Fall 2009, Sllide 194


 * Sources:**
 * Analyzing Multivariate Data, by James Lattin, Douglas Carroll and Green ([|Amazon]), page 44
 * EMSE 271, Fall 2009