SkipCaptcha
35,519
edits
Changes
Jump to navigation
Jump to search
Line 2:
Line 2: + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Essay:Analyses of misleading polls (view source)
Revision as of 18:45, October 26, 2020
, 18:45, October 26, 2020added a section
__TOC__
__TOC__
=October 26, 2020=
{{Essay By|Liberaltears}}
[https://drive.google.com/file/d/1jZ9sY9pNj52MdJPsN9MW4KHKvDsjNC9c/view This CBS/YouGov poll] conducted for the 2020 presidential and Senate elections in [[Georgia]] was done in late October 2020, stating that out of a sample of 1,090 ''likely voters'', Biden and Trump are tied in the presidential election and that Sen. [[David Perdue]] is leading [[Democrat]] challenger [[Jon Ossoff]] by one percentage point; the overall margin of error is ±3.4%. Specifics need to be examined.
It's firstly important to observe the crosstabs for the presidential/Senate election preferences by party identification. According to them, both Trump and Perdue are leading their Democrat challengers among [[independent voters]] by ten percentage points, which contradicts the liberal narrative promoted by less trustworthy polls that independents supposedly are breaking from the [[Republican Party]]. Out of 1,083 people surveyed just the Senate race, 388 were Republicans, 394 were Democrats, and 290 were independents. For the presidential race where 1,084 were sampled, 389 were Republicans, 391 were Democrats, and 293 were independents. As the following tables show:
{|class="wikitable" style="float:left;"
!
![[Republican]]
![[Democrat]]
![[Independent voters|Independent]]
!Total
|-
|'''[[Donald Trump]]
|95%
|2%
|50%
|48%
|-
|'''[[Joe Biden]]
|3%
|97%
|40%
|48%
|-
|'''Total
|98%
|99%
|90%
|96%
|}
{|class="wikitable" style="float:left;"
!
![[Republican]]
![[Democrat]]
![[Independent voters|Independent]]
!Total
|-
|'''[[David Perdue]]
|92%
|3%
|50%
|47%
|-
|[[Jon Ossoff]]
|3%
|93%
|40%
|46%
|-
|'''Total
|95%
|96%
|90%
|93%
|}
Since [[independent voters]] appear to favor [[Donald Trump]] and [[David Perdue]] by clear margins, the only way they could be tied in their respective races is if Democrats are over-weighed and/or independents are under-weighed in the poll. Based off the previous numbers listed for the Senate race in terms of party affiliation, 35.83% were [[Republicans]], 36.38% were [[Democrats]], and 26.78% were [[independent voters|independents]]. For the numbers for the presidential race, 35.78% were Republicans, 36.24% were Democrats, and 26.88% were independents. Are they an accurate model to use for the 2020 elections in [[Georgia]]? Assuming that the turnout rate (and thus rates by party) will be similar and comparable to the [[2016 U.S. presidential election]], an exit poll from the previous cycle can be used to compare. According to this [https://www.cnn.com/election/2016/results/exit-polls/georgia/president 2016 CNN exit poll] for the presidential election then, 36% were Republicans, 34% were Democrats, and 30% were independents. This means that the [[CBS]]/YouGov poll, when compared to the exit poll, under-weighed independents and slightly over-weighed Democrats. And to weigh properly:
<pre style="width:33em">
Trump: (36% × 95%) + (34% × 2%) + (30% × 50%) = 49.88%
Biden: (36% × 3%) + (34% × 97%) + (30% × 40%) = 46.06%
Perdue: (36% × 92%) + (34% × 3%) + (30% × 50%) = 49.14%
Ossoff: (36% × 3%) + (34% × 93%) + (30% × 40%) = 44.70%
</pre>
It becomes clear now that Trump and Perdue hold respectively a 3.82% lead and a 4.44% lead over their [[Democrat]] challengers. While this calculation may not be the most accurate way to predict the 2020 elections, it helps shed light on the [[liberal bias]] in the poll's over-weighing of Democrats and under-weighing of independents to "conclude" a dead heat in [[Georgia]] when Republicans are most likely inclined to win the state. It's also important to note that while [[MSM]] sources may mention increasing urban turnout to potentially flip the state blue, Trump has gained strong support from blacks, [https://cnsnews.com/article/national/andrew-davenport/rasmussen-trump-approval-rating-36-among-black-voters] [https://www.washingtonexaminer.com/politics/trumps-approval-rating-among-black-voters-jumps-nearly-60-during-gop-convention] [https://onenewsnow.com/politics-govt/2020/10/24/rasmussen-poll-blacks-give-trump-46-approval-rating] which may likely secure his chances of winning overall.
=October 25, 2020=
=October 25, 2020=