This page is part of an undergraduate assignment at Davidson College.
| Works Cited |
Results
Predictions Discussed and Overall Trends
First, I will talk about my predictions. Well, I was wrong
about John Kerry’s national politician variable being significant,
yet I was right that he had a very good model, with the most significant
variable out of any of the models. However, I said that he wouldn’t
have many significant variables so I was incorrect, but perhaps that is Timothy
Hill’s fault who caused me to believe in the “glass ceiling” with
respect to endorsements. Dean was surprising that despite his train wreck
at Iowa, he still had four significant variables with a very representative
model. My Gephardt prediction was probably not well thought out because there
was lots of talk before Iowa that he was going to drop out if he didn’t
win the state. Other than that, I am very happy with my Edwards underdog
prediction with the only one with national politician significance.
Other than my discussing my predictions, I do not discuss Mondays’ significance
completely. Due to news cycles being cyclical it makes sense that one-day is
more significant across the board than others. What this really means is that
the majority of the news comes out on Tuesday, which makes sense because most
of the press releases go out on Monday after the weekend.
The R-squared values were so impressive that either, we weren’t doing
enough with the model, or most likely there was a erogeneity error, which I
will go into a little bit in the next section.
© Davidson College, 2004, Department of Political
Science, Davidson College, Davidson, NC 28035
Send comments, questions, and suggestions to Dr.
Patrick Sellers, Professor of Political Science.
Created: 4/29/2004. Last updated: 5/03/2004.