X users call out CNN for ‘oversampling’ Republican voters in poll to set ‘anti-Biden narrative’
CNN Poll: Biden's job ratings low, voters favor Republican nominee for 2024 election. Critics accuse CNN of being bias.
A recent CNN poll has come under fire from several X users who have accused the outlet of "oversampling" Republicans in a survey ahead of the 2022 midterms.
The poll, which was conducted by SSRS and released on Wednesday, showed that Republicans had a narrow edge over Democrats on a generic ballot question among likely voters, 51% to 47%.
However, a paper on the survey revealed that CNN had modified the poll's sample, including a higher number of Republicans and a reduced count of Democrats compared to the actual population.
According to the American news outlet, the poll’s sample was “weighted to match the partisan makeup of the electorate in the last midterm election, when Republicans had a slight advantage in turnout”.
In its, defence CNN claimed that this was done to “reflect the historical pattern of lower Democratic turnout in midterm years”.
However, critics pointed out that this was a “flawed and biased assumption”, as there is no guarantee that the 2022 electorate will resemble the 2018 one, especially given the changing demographics and political dynamics of the country.
Some of the critics were prominent voices on X, such as Palmer Report and Mueller She Wrote, who accused CNN of “rigging” its own poll to create a false narrative of a Republican wave. They argued that CNN was trying to influence public opinion and voter behavior by creating a self-fulfilling prophecy of Democratic defeat. Echoing Republicans, they also questioned CNN’s credibility and journalistic ethics, and urged Democrats not to be discouraged by the poll results.
Palmer Report tweeted, “CNN is now admitting that it rigged its own poll to produce a fake result so it could stir up ratings. At this point every major political news outlet is basically the National Enquirer.”
Mueller, she wrote tweeted, “All I’m saying is read the fine print.”
In his tweet, Rich Baris another verified X user asserted that there are ample grounds to disregard CNN.
"Plenty of reasons to dismiss the CNN Poll conducted by SSRS for being an outlier all cycle and for this ridiculous result published today.
But "oversampling" is not the same as 'overrepresentation'. Republicans were 'oversampled' for the primary, but weighted down for general."
Another X user accused the outlet of fabricating their own news, and refuted their work ethics.
“CNN desperately wanted an anti Biden narrative so what did they do? They commissioned a poll comprised of 59.7% Republicans(!) to generate a predictably bad outcome for Biden and have now blabbed about it nonstop all day. "You literally can't make this **** up. They are literally making their own news," the tweet read.
On the other hand, President Joe Biden continues to grapple with persistently unfavorable job ratings, widespread apprehensions about his age, and diminishing confidence among voters aligned with the Democratic party, as revealed in a recent CNN Poll conducted by SSRS.
However, a Community Note points out that this is just a sampling method and not an error.
“This poll followed standard polling techniques. Oversampling is a statistical technique for obtaining more precise information about particular subgroups (in this case, Republican primary voters). The final sample is reweighted so that poll results are not biased.”
An article in Pew Survey explained: "Oversampling is the practice of selecting respondents so that some groups make up a larger share of the survey sample than they do in the population. Oversampling small groups can be difficult and costly, but it allows polls to shed light on groups that would otherwise be too small to report on."
It adds: "This might sound like it would make the survey unrepresentative, but pollsters correct this through weighting. With weighting, groups that were oversampled are brought back in line with their actual share of the population – removing the potential for bias. When people think about opinion polls, they might envision taking a random sample of all adults in the U.S. where everybody has the same chance of being selected. When selected this way, the sample on average will look just like the full population in terms of the share that belongs to different groups."
ALSO READ| Nikki Haley emerges as sole Republican who can beat president Joe Biden in 2024, says CNN poll
In a hypothetical rematch between Biden and former President Donald Trump, no distinct frontrunner emerges, despite Trump's substantial lead in the GOP primary race.
Almost half of registered voters (46%) express the belief that any Republican presidential nominee would be a superior choice to Biden for the 2024 election.
