The methodology for conducting the scientific survey of Nebraska opinion has changed from past World-Herald Polls.
In consultation with Wiese Research Associates, we determined it was worth the added effort and expense to include cellphones in the telephone survey.
Wiese Research, which has conducted public opinion surveys for The World-Herald for several years, also used the statewide voter registration list this year as the basis for the random survey.
The combined addition of cellphones and the registration list provides the highest probability that every registered voter had an equal chance of participating in the survey.
With nearly 30 percent of all households now relying exclusively on cellphones, continuing to call only land line numbers increased the risk of inaccuracy to what we considered unacceptable levels. We recognize that many reputable pollsters continue to limit their calls to land lines.
The survey of 800 registered voters was conducted Sept. 17-20. The maximum expected statistical margin of error for all statewide results is plus or minus 3.5 percentage points.
The respondents were 400 registered voters from the 2nd Congressional District and 200 each from the 1st and 3rd Districts. The 2nd District was oversampled to provide an acceptable margin of error of 4.9 percentage points for the results in the presidential and U.S. House race in that district.
A weighting system was used to correct for the oversampling and give each of Nebraska's three congressional districts equal voice in the statewide result totals. Quotas were utilized to make sure age, income, gender, partisan and geographic areas were appropriately represented according to their actual proportions in the registered voter population.
As with all scientific surveys, The World-Herald Poll should not be used to predict the outcome of any election. The survey attempts to capture a snapshot of voter sentiment.