Shame on you, Zogby... You knowingly or unknowingly participated in a DNC scam to influence the election in favor of Kerry by demoralizing potential voters in later time zones and therefore attempting to depress voting.
Exit Polls Come Under Post-Election Criticism
Reuters | Wed Nov 3, 2004 07:30 PM ET | Steve Gorman and Arthur Spiegelman
LOS ANGELES (Reuters) - Exit polling, long a key indicator of election trends, emerged on Wednesday as a big loser in the 2004 election after Internet pundits latched onto early data to mistakenly report a commanding lead for Democrat John Kerry in key states.
The use of the raw, preliminary data on freewheeling Web sites triggered a brief but doomed wave of euphoria among Kerry supporters and a sharp drop on Wall Street. The data also wound up being telegraphed by mainstream media in TV reports.
The truth was that in the end Kerry did not win states like Ohio and Florida, where early exit polls showed him ahead by several percentage points.
On the day after the elections, experts and pollsters tried to figure out what went wrong -- were the exit polls off base or was the data misused?
"We have an exit poll crisis," said pollster John Zogby, whose own Election Day predictions that Kerry would take Ohio and Florida proved wrong. He said he had used the exit poll data to confirm conclusions he had based on his own polling.
Zogby, in an interview with Reuters, for whom he conducts polls, added, "As polling professionals, we test and judge our samples based on exit polls.
"The first and second rounds yesterday were not good. It used to be that whatever came out in the first round generally held, but yesterday the first round made it look like a Kerry landslide was in the works and he was sweeping states he didn't win, and that is troublesome."
The origin of at least some online reports giving Kerry a hefty lead in key battleground states was data collected by pollsters Edison Media Research and Mitofsky International for the National Election Pool, a consortium of TV networks and other news outlets.
"Some Web sites had data that claimed to be ours. Some of it may have been ours, some of it may have been made-up data, some of it may have been misinterpreted data," said Joe Lenski, vice president of Edison Research.
"A lot of this is whispered-down-the-line stuff, where someone grabs one piece of data and sends it to someone else."
Lenski said early results were more favorable to Kerry than they were for Bush, owing in part to the fact that more women then men tend to vote early in an election and that women in this race favored Kerry.
But he said the size of the Kerry leads attributed to exit polling were exaggerated by online political pundits known as bloggers, who also appeared to have ignored the caveats and analysis supplied with the data. "You can tell that anything that appeared online was not interpreted by anyone that had a sophisticated understanding of the workings and limitations of early exit polling," he said.
Early exit polling can be by its nature heavily skewed to one side or another and is not intended to be used to project races or be reported to the public, especially before polls close, Lenski said.
Rather, it is used by news outlets as early guide posts to possible trends that may unfold later.
While election watchers credited the TV networks with steering clear of early exit poll number in their own reporting, they noticed that partisans of the Kerry camp referred to those results in their on-air commentary, helping propagate the flawed information in the mainstream media.
Reuters/VNU
Exit Polls Come Under Post-Election Criticism
Reuters | Wed Nov 3, 2004 07:30 PM ET | Steve Gorman and Arthur Spiegelman
LOS ANGELES (Reuters) - Exit polling, long a key indicator of election trends, emerged on Wednesday as a big loser in the 2004 election after Internet pundits latched onto early data to mistakenly report a commanding lead for Democrat John Kerry in key states.
The use of the raw, preliminary data on freewheeling Web sites triggered a brief but doomed wave of euphoria among Kerry supporters and a sharp drop on Wall Street. The data also wound up being telegraphed by mainstream media in TV reports.
The truth was that in the end Kerry did not win states like Ohio and Florida, where early exit polls showed him ahead by several percentage points.
On the day after the elections, experts and pollsters tried to figure out what went wrong -- were the exit polls off base or was the data misused?
"We have an exit poll crisis," said pollster John Zogby, whose own Election Day predictions that Kerry would take Ohio and Florida proved wrong. He said he had used the exit poll data to confirm conclusions he had based on his own polling.
Zogby, in an interview with Reuters, for whom he conducts polls, added, "As polling professionals, we test and judge our samples based on exit polls.
"The first and second rounds yesterday were not good. It used to be that whatever came out in the first round generally held, but yesterday the first round made it look like a Kerry landslide was in the works and he was sweeping states he didn't win, and that is troublesome."
The origin of at least some online reports giving Kerry a hefty lead in key battleground states was data collected by pollsters Edison Media Research and Mitofsky International for the National Election Pool, a consortium of TV networks and other news outlets.
"Some Web sites had data that claimed to be ours. Some of it may have been ours, some of it may have been made-up data, some of it may have been misinterpreted data," said Joe Lenski, vice president of Edison Research.
"A lot of this is whispered-down-the-line stuff, where someone grabs one piece of data and sends it to someone else."
Lenski said early results were more favorable to Kerry than they were for Bush, owing in part to the fact that more women then men tend to vote early in an election and that women in this race favored Kerry.
But he said the size of the Kerry leads attributed to exit polling were exaggerated by online political pundits known as bloggers, who also appeared to have ignored the caveats and analysis supplied with the data. "You can tell that anything that appeared online was not interpreted by anyone that had a sophisticated understanding of the workings and limitations of early exit polling," he said.
Early exit polling can be by its nature heavily skewed to one side or another and is not intended to be used to project races or be reported to the public, especially before polls close, Lenski said.
Rather, it is used by news outlets as early guide posts to possible trends that may unfold later.
While election watchers credited the TV networks with steering clear of early exit poll number in their own reporting, they noticed that partisans of the Kerry camp referred to those results in their on-air commentary, helping propagate the flawed information in the mainstream media.
Reuters/VNU