top of page

Why This?

As we near voting season for 2024, there has been a large amount of online political tension in the past four years, which I hope translates to more people voting this year. To understand if this logic made sense, I turned to 2020 to see if there was a significant increase in voter turnout over 2016.

Creating a Hypothesis

To do hypothesis testing, we need a hypothesis (Hₐ) and a null hypothesis (H₀)

My hypothesis (Hₐ) is that there was an increase in voter turnout in 2020 over 2016. As a result, the null hypothesis (H₀) is that voter turnout in 2020 was the same as in 2016.

We also have a threshold for data significance (⍺) of 0.05

The Data

I scoured the internet for voter turnout, and after days of work found the US Election Project. A site dedicated to tracking The United States of America's elections. I then compiled it into what you see below.

If you are having trouble viewing the spreadsheet, you can access it here.

What Does This Mean?

To understand if there is significant data to support this hypothesis, we must first find the z-score. The z-score is a marker on a spectrum of primarily -3.4 to 3.4. This can be translated into a p-value, to indicate on a scale of 0 to 1 where this data stands.

Let's Calculate It

To get the z-score, we use the formula:

Where x is the sample mean (2020), μ is the population mean (2016), σ is the standard deviation, and n is the sample size. Filling in these values results in the following results:

This can be rounded to 6.74, which if we look at a p-value chart is above the values of the chart, which means that the p-value is about 1. However, because the z-value is positive, the data is right-tailed, and we must subtract the p-value from 1. This is 1-~1, which is equal to about 0

Conclusion

This p-value is less than 0.01 (right tailed), so we can conclude that there is highly significant evidence to back up my hypothesis. Therefore, we reject the null hypothesis.

Let's Speculate

The Determinant: Voter Turnout

U.S. Presidential Elections

bottom of page