Section 3.6
Problems Involving Three Planes
October 25, 2007
Suppose we have three distinct planes. Can we determine how they
intersect without calculating the point(s) of intersection?
We examine their normal vectors for:
collinearity
coplanarity
and exa

where dfH is the degrees of freedom asociated with the reduced linear model or the linear
model under the null hypothesis and dfU is the degrees of freedom associated with the full
model. Usually, dfH = number of independent parameters estimated in the re

LECTURE NOTES: STATISTICS 6520
(LINEAR MODELS - Fall 2003)
Quadratic Forms
Let x = (x1 , x2 , . . . , xn )T be a n 1 vector and A = (aij ) an n n symmetric matrix. In
our discussions, we will encounter expressions of the form
T
Q(x) = x Ax =
n
X
=
n n
XX

MEMORIAL UNIVERSITY OF NEWFOUNDLAND
DEPARTMENT OF MATHEMATICS AND STATISTICS
SAMPLE MIDTERM EXAM - STATISTICS 6520 FALL 2011
Instructor: A. Oyet
Date: November 1, 2001
Name(Surname First):
Student Number
INSTRUCTIONS
1. Time allowed 75 minutes.
2. Attempt

STATISTICS 6520 - LINEAR MODELS
ASSIGNMENT 3, DUE November 14
All questions refer to the full rank model
y = X + , rank(X) = p < n, Nn (0, 2 In ).
1. Let H be the estimator of the parameter vector under the null hypothesis H0 : C = 0
and be the unconstrai

STATISTICS 6520 - LINEAR MODELS (FALL 2013)
ASSIGNMENT 2, DUE October 31
All questions refer to the full rank model
y = X + , rank(X) = p < n, Nn (0, 2 In ),
whenever a model is not specied.
1. Suppose that E[Y] = , A = 0 and cov[Y] = 2 V, where Aqn is a