Summary of “The Assumptions of the Linear Regression Model,” page 145148
Poole and O’Farrell begin their article by first discussing a history of regression modeling
in the fields of geography and planning which disregard sufficient attention the assumptions of
the linear regression model that are necessary to ensure that the models results are valid.
They
claim that this practice has stemmed from a common problem amongst writers in these fields.
That is, they discuss regression analysis too briefly so as to not allow sufficient discussion of the
assumptions.
Those who did broach the subject of the regression assumptions, like Cole and
King in 1968, did so insufficiently as they only mention one of the models assumptions.
In
1968, Yeate’s published a work which improved with 3 assumption references.
Then LJ King, in
1969, alluded to all seven assumptions yet he did so in an inaccurate and ambiguous manner.
Similarly, Colenutt also wrote with regard to all seven assumptions.
Though King was
addressing geographers and Colenutt addressing planners, they share a similarity in that they
both felt it necessary to warn their colleagues of the necessity to deal with the assumptions
properly, yet both of them describe how to do so inadequately.
After discussing this history of errant regression analysis in the relevant fields, Poole and
Farrell transition to a review of the assumptions of linear models.
Prior to addressing the
assumptions, the model is first summarized.
Poole and Farrell define the linear regression
model as Y = a + [sum of (biXi)] + u where Y is the dependant variable; X1,…,Xi are i
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
This is the end of the preview.
Sign up
to
access the rest of the document.
 Spring '10
 RCOLLIER
 Regression Analysis, conditional mean, ui given Xi

Click to edit the document details