UpDate - Vol. 12, No. 5, Page 6
October 1, 1992
Polling voters lets students apply stats to 'real world'
Move over, Mr. Gallup. Take a back seat, Mr. Harris. Delaware
high school student pollsters will soon be on the telephone, surveying
voters about their choices for state and national elections and
predicting the winners for Election Day 1992.
Scheduled for Oct. 15-19, the 1992 Voter Preference Survey,
coordinated by the University, is the third student-powered poll
during a presidential race in Delaware. The 1984 and 1988 surveys were
mostly on target, accurately predicting the winning candidates,
according to John H. Schuenemeyer, professor of mathematical sciences,
one of the originators of the project.
He and Kathleen Hollowell, coordinator in the Mathematical
Sciences Teaching and Learning Center, have conducted workshops for 39
volunteer teachers from across the state. They represent 15 out of the
state's 19 public school districts, the secondary schools of the
Wilmington Catholic diocese and one private school.
Approximately 1,000 students will participate, and it is
anticipated that each will successfully complete three to five
surveys. The number of participating students has increased with each
survey.
The callers will ask likely voters whom they plan to support in
this election, from the office of state insurance commissioner on up
to president. Another question asks what voters consider to be the
most important national issue in 1992. Voters also will be asked about
their party affiliation, sex, age group and ethnic origin.
"We hope the public will cooperate with the students involved in
the project. It is an educational experience for them and also
performs a public service. Individual responses are all confidential,"
Schuenemeyer said.
The statewide results will be tallied immediately by members of
the Delaware chapter of the American Statistical Association, and the
data will be sent to classrooms for analysis.
"The project teaches students the purpose and use of statistics,
how statistics are gathered and analyzed mathematically. We welcome
the opportunity to work with teachers on a timely project that
demonstrates to students the application of statistics to the "real
world." We are pleased with the teachers' response and participation,"
Hollowell said.
Other lessons also are learned, Schuenemeyer said. Surveying
voters interests the students in both the election and the issues, and
makes them more aware of the political process. The training on
interviewing techniques and the actual interaction with the public
also are beneficial, he said.
"When doing a poll of this kind, you have to survey the right
people-that is a random sample of people who are Delaware citizens of
voting age who intend to vote.
"Sometimes, an event happens while the poll is being conducted or
afterward that changes the voters' minds. In one case, there was a
debate between Thomas Carper and Elise du Pont that appeared to alter
the outcome of the election, so that while the winner was correctly
predicted, the percentage was not as accurate as in the other races,"
Schuenemeyer said.
The 1992 Voter Preference Survey is a project of the Delaware
chapter of the American Statistical Association; the University's
Mathematical Sciences Teaching and Learning Center and Department of
Mathematical Sciences; and the Delaware Secondary Schools.
In addition to Hollowell and Schuenemeyer, James Lucas, a
statistician, formerly with the DuPont Co. and now with Drexel
University, and Robert Gorman, a mathematics graduate student, also
are working on the project.
-Sue Swyers Moncure