This polling average was created using data from the FiveThirtyEight polling github page. The polls were weighted by pollster rating, poll methodology (eg online vs. live phone), and the population that was polled (likely voters, registered voters, all voters). An exponential decay formula was also applied to smooth out the data, the idea being that a raw polling average is overly aggressive at detecting minor and fleeting fluctuations. The exponential decay function eliminates some of this noise and provides a better idea of overall trends.
iframe {
min-width: 100%;
}
iFrameResize({
heightCalculationMethod: 'taggedElement'
});
This is a visualization comparing the prices listed on the horserace gambling website PredictIt versus candidate polling in the 2020 Democratic primary. The gist of what I’m interested in is how closely PredictIt’s market offers follow candidate polling averages. Keep in mind that the PredictIt market is trying to measure probability of a candidate’s ultimate nomination, not their current polling average.
This is an analysis of 205 emails sent out from July 2019 to September 2020. A detailed explanation of the methodology and dataset can be found in the Discussion portion of this doc.
Takeaways
Weekdays are better than weekends
Emails sent before 9am don’t perform very well
Take advantage of MailChimp’s “Send with TimeWarp,” especially for morning emails
For late afternoons, 5pm is better than 6pm
What’s the best day to send emails?
There is a belief in the Political Twittersphere, especially among Bernie supporters, that phone polls are biased towards candidates who perform better among older constituencies (primarily Biden) due to their reliance on landlines. Live phone methods do tend to undersample younger voters because they rely heavily on landlines. However, most pollsters worth their salt will account for this representation discrepancy through their weighting processes.
To get at the question of differences in polling results based on methodology, I’m going to look at some polling data provided by 538 to create separate plots for the top eight candidates looking at their poll percentages over time and broken out by methodology.
Introduction
Study Design
A study on congressional candidate fundraising and expenditures in the 2018 midterm elections is considered. The variables are: Candidate Contribution (CAND_CONTRIB), Candidate Loans (CAND_LOANS), Other Loans (OTHER_LOANS), Total Contributions from Individuals (TTL_INDIV_CONTRIB), Total Contributions from Political Party Organizations (POL_PTY_CONTRIB), and Contributions from Other Political Organizations (OTHER_POL_CMTE_CONTRIB), all recorded as dollars. Total Spending in the race was also controlled as a proportion of a candidate’s spending out of the total spending of the top two candidates in the race (PROPORTION_SPENT), and the candidate’s incumbency status and party were also included (incumbent, party).