Home > Archives > Election Insider
July 29, 2010 Content provided by DDC Advocacy
Those who observe and analyze political elections pay a great deal of attention to survey research. Sometimes, a reality jolt is necessary to remain cognizant that not all polls are equally accurate. This week's Oklahoma Democratic gubernatorial primary illustrates some glaring polling inaccuracies. Two separate studies, the Cole Hargrove Snodgrass consulting firm's statewide Sooner Poll, which the Oklahoma media heavily relies upon, and their follow-up Sooner Survey pulled separate samples for the same race, and found similar results. They both predicted, with less than a week remaining in the primary battle, that Attorney General Drew Edmondson would handily defeat Lt. Governor Jari Askins. The Sooner Poll (7/16-21; 755 OK Democratic primary voters) predicted a 49-33% Edmondson advantage. The CHS Sooner Survey (7/18-20; 400 likely Democratic OK voters) gave the AG an eleven point lead, 38-27%. The results: Askins won 50.3% to 49.7%, a margin of just 1,493 votes, but far ahead of her projected finish. On the other hand, the Sooner Poll did predict that GOP Rep. Mary Fallin would win her party's nomination outright against three opponents and score in the neighborhood of 56%. They pinpointed almost the exact results as Fallin claimed victory with 55%. As Oklahoma reminds us, polling can never be considered an exact science.