# Margin of Error for Statistical Predictions

Emrullah Yiğit

Mehmet Efe Zengin

What would you think if I were to say that a poll company made 100% correct predictions? How can someone know about the elections or the effectiveness of a vaccine? These are all questions that the statistics -a math discipline- considers. Always, there is an uncertainty; however, it is still possible to make accurate predictions as long as the error or uncertainty is understood. For example, when the effectiveness of the vaccine for COVID-19 is tested, people could vaccine everyone to put an exact result. No! That’s never possible to access 100% of the population in the real world. However, you can choose a sample that can represent most of the population, such as ethnic diversity, gender, age… In the case of testing the vaccine on 10 people and finding 7 of them effective, that could say 70% effectiveness. However, in statistics, it is called margin of error that creates a confidence interval that the researcher is confident that the interval will capture the true proportion of the population. More the size of sample, less the margin of error and more accurate the confidence interval. Considering the accuracy that you would like to reach, you choose more to test on as much as you can.

References:

Harel, Ofer. “What Is a Margin of Error? This Statistical Tool Can Help You Understand Vaccine Trials and Political Polling.” Phys.org, Phys.org, 7 Jan. 2021, phys.org/news/2021-01-margin-error-statistical-tool-vaccine.html.