Reply #45: Hello Helderheid

[jkd]

Fellow Utahan, I sent this communication to Dr. Freeman:

Dr. Freeman,
Why do you continue to cite the BYU exit polls as evidence of exit poll accuracy. Mifofsky projected Kerry at 30.5% and the students had Kerry at 26.5% in Utah. The Utah poll missed by .5% and the NEP was off by 4.5%. These data seem to contradict your suggestion that Mitofsky must also be accurate because they are professionals and not just students.

This was Mr. Freeman's response:


"Good point; I thought about your point when I wrote it, but it's next to impossible to reconcile every piece of data ... What do you conclude from this? Do you have a particular change that you suggest for the paper or it's conclusions?"

The Utah exit polls were far superior to Mitofsky because they were designed to predict the election unlike the NEP polls. They used 900 students to poll 90 precincts. Mitofsky only polled 15 precincts. The Utah polls were right on and NEP missed by 4.5%. TIA assumes all exit polls are created equal; they are not. If Mitofsky had properly conducted polls to predict election results, then all this analysis might mean something. What might this suggest about Mitofsky's polls in other states besides Utah.

I presented more information about the BYU polls to Dr. Freeman, but he has not chosen to answer my last reply.