Election Polling Project

Summary

In the Election Poll Project, I utilized R programming to conduct a comprehensive analysis of polling data related to the US presidential elections in 2016 and 2020. The focus of the project was on exploring and comparing polls in specific states and evaluating changes in polling outcomes over the two election years.

Project Details

Data Exploration

• Data Cleaning: Utilized R code to thoroughly explore and clean polling data, ensuring accuracy and reliability for analysis.
• State and Time Selection: Organized the data by months and focused on specific states for detailed analysis.

Geographical Analysis

• State Focus: Concentrated on states like Michigan, Georgia, and North Carolina to explore polling trends for the 2016 and 2020 elections.
• State Insights: Analyzed the nuances of polling trends within these states to identify changes or consistencies in voter behavior.

Temporal Comparisons

• Yearly Comparisons: Conducted a temporal analysis to compare polling data across 2016 and 2020, identifying shifts or patterns in voter preferences.
• Statistical Tests: Used paired t-tests and Wilcoxon signed-rank tests to calculate p-values and determine polling accuracy.
• Poll Accuracy: Compared the polling data predictions with actual election results, examining sampling bias and response bias.

State-to-State Comparisons

• Iowa and Florida: Extended the analysis to compare polling trends in Iowa and Florida, exploring changes or similarities in these states over the two elections.
• Prediction Accuracy: Evaluated how well the polls predicted the outcomes in these battleground states.

Overall Differences

• National Trends: Visualized the overall polling differences between the 2016 and 2020 elections, highlighting key battleground states.
• Percentage Differences: Generated maps to visualize percentage differences and shifts in voter preferences across the country.
• Electoral Vote Shifts: Analyzed changes in electoral votes by comparing percentage shifts between leftist and rightist voter bases from 2016 to 2020.

Report Generation

• Comprehensive Report: Compiled the findings into a detailed report, discussing insights derived from polling data and providing suggestions for reducing bias in future polls.

Conclusion

This project demonstrated my proficiency in R programming and my ability to analyze complex political datasets. The Election Poll Project reflects my capability to perform data-driven analysis and extract meaningful insights from election polling data, offering valuable perspectives on polling accuracy and voter trends across election years.