Stephen suggested that a design flaw in some of those Florida ballots caused
the problem rather than an inability to follow directions. You can take a
look at the ballot yourselves at
http://dailynews.yahoo.com/h/p/nm/20001108/pl/mdf129505.html
This ballot looks like a great example of how not to design a fixed choice
instrument. We would not let our students use this design for research
purposes. At this time there seems to be a legitimate question about whether
this ballot was designed in error. There is nothing like a good conspiracy
theory to keep things interesting. OR we could use this opportunity to build
enrollment in our research methods classes.
For those of us teaching statistics, there is another teaching opportunity
here. With 19,000 votes lost when much less then 2,000 make the difference
in the current count it is no wonder that the polling folks were wrong. I
wish that the other polling problems had been identified because they too
are relevant to that issue.
Dennis
Dennis M. Goff
Dept. of Psychology
Randolph-Macon Woman's College
Lynchburg, VA 24503
