How the polling was conducted

The Columbus Dispatch – May 23, 2004

The poll by Saperstein Associates for The Dispatch and WBNS-TV (Channel 10) was based on telephone interviews with 403 randomly selected adults in the Columbus area from Monday through Thursday.

The Columbus polling company uses a random-digit dialing technique to contact potential poll participants. In all, 9,647 calls were made to 5,525 numbers from Columbus telephone exchanges to obtain the 403 interviews.

The results were weighted to ensure that the sample accurately reflected the adult population in the Columbus area.

The margin of sampling error is plus or minus 4.9 percentage points. In theory, that means in 19 cases out of 20, the results obtained in this survey will differ by no more than 4.9 percentage points in either direction from the results that would have been obtained by interviewing all adults in the Columbus area.

The error margin is larger for subgroups of respondents.

In addition to sampling error, all surveys are subject to other potential sources of error, such as bias in the wording of questions, deficiencies in the quality of the interviews or other types of imprecision.