Essay:An analysis of another misleading poll

This essay is an original work by Liberaltears. Please comment only on the talk page.

(see here for mathematical calculations) See also: Essay:An analysis of a misleading poll

This poll by Redfield & Wilson was published on June 25, 2020 covering the 2020 U.S. presidential election and Senate elections in key swing states. For the 2020 U.S. Senate special election in Arizona, the poll "concluded" that Republican interim senator Martha McSally was trailing likely Democrat opponent Mark Kelly by 15 percentage points. Having a sample size of only 865 respondents, there are specifics that need to be analyzed; two-way tables are given for sample makeups in categories for gender, age, region, 2016 vote in the presidential election, and intended vote in the 2020 U.S. presidential election. Of these, the gender and region data will be thoroughly assessed.

Analyzing gender sample makeup/conclusions

(see here for mathematical calculations regarding this section)

The following is a table that shows preferences between Sen. McSally and Kelly in the Senate election by gender:

	Male	Female
Martha McSally	34%	34%
Mark Kelly	52%	46%
Other	2%	4%
Don't Know	13%	16%
Total	101%	100%

While the sample makeup is mostly weighed properly (exactly 50% male and 50% female), the real problem here is that the poll is asserting that men are somehow going to overall vote to the left of women by 6 percentage points (in net terms), a mostly absurd notion. In Arizona (as shown in exit polls), men voted to the right of women by a net amount of 17 points in the 2016 presidential election, 11 points in the 2016 Senate election, 4 to 19 points in the 2018 Senate election, and 6 to 15 points in the 2018 gubernatorial election. Furthermore, the entire sample appears to be very left-leaning, making the poll conclusions highly questionable.

Analyzing region sample makeup/conclusions

The following is a table assessing respondents' preferences between McSally and Kelly by their region in Arizona:

	Maricopa County	Pima County	Other Counties
Martha McSally	31%	40%	37%
Mark Kelly	51%	53%	40%
Other	2%	2%	6%
Don't Know	15%	4%	18%
Total	99%	99%	101%

In finding the sample makeup by region, the "Don't Know" section can be ignored. This would then leave three individual rows and columns. With such, a three-variable system can be set up and solved, with the calculated results for sample makeup roughly being 62.15% of respondents from Maricopa County, 14.09% from Pima County, and 24.59% from all other counties (note that this does not add up to exactly 100%). While this does seem to be weighted properly (for the most part), the real problem is that, similar to the gender makeup, the conclusions are absurd. According to the poll, Maricopa County will supposedly vote to the left of Pima County by a net of 7 percentage points, a mostly absurd notion. Pima County had voted to the left of Maricopa County by around 11.4% in the 2018 Senate election, 16.8% in the 2016 presidential election, 18.8% in the 2016 Senate election, 17.8% in the 2012 presidential election, 20.2% in the 2012 Senate election, etc. While it's true that many of the neocons in Maricopa County who supported establishment RINO John McCain and currently oppose Donald Trump are likely to turn against McSally (as McCain donors are), it's unlikely that such numbers will occur. In addition, it's unlikely that such numbers will be the case for the rest of the Arizona counties (that lean Republican), as the entirety of sample appears to be skewed to the left.

Final points

Since the poll overall has appeared to have weighed several groups accurately though selected a left-leaning sample, there would be no practical use to further weigh the poll. Furthermore, because the entire sample is left-leaning, the poll should be distrusted.