The History and Challenges of Election Polling
TLDR Election polling has a long history, starting with informal straw polls and evolving into more formal methods like door-to-door and telephone surveys. However, the decrease in phone polling response rates has led to the use of alternative methods like robo calls and online polls. While polls can be manipulated, legitimate polling groups use weighting and additional variables to ensure a representative sample and account for a margin of error. The 2016 election highlighted the need for more accurate and nuanced presentation of election predictions, and internet polling is a new and disruptive form of polling that is still being figured out.
Timestamped Summary
00:00
Election polling is often misunderstood and misrepresented by the media, leading to misconceptions about its accuracy and reliability.
04:45
Election polls in 2016 were not significantly off from historical averages, and while state polling had a slightly higher error rate, overall polling was not inherently flawed or inaccurate, despite media portrayal.
09:33
The history of polling dates back to informal straw polls, but it wasn't until the 1930s that more formal polling methods were established, starting with Gallup's door-to-door surveys, which eventually transitioned to telephone surveys.
14:08
Phone polling response rates have significantly decreased over the years, making it more expensive and difficult to obtain a sufficient number of survey responses, leading to the use of robo calls and online polls as alternative methods.
19:02
Phone polling involves calling individuals repeatedly to ensure a diverse range of respondents and improve the accuracy of the results by weighting the responses to match real-world demographics.
23:32
The margin of error in polling allows for a 95% confidence interval, meaning that almost all results will fall within a few percentage points of the middle, providing a snapshot of public opinion at a given moment.
28:05
Polls can be manipulated to present a certain narrative, but legitimate polling groups use weighting and additional variables to ensure a representative sample and account for a margin of error.
32:57
In 2016, the state polling was off because pollsters did not account for the large number of non-educated white people who came out to vote for Donald Trump.
38:10
The 2016 election showed that pollsters failed to account for the large number of non-college educated white voters who came out to vote for Trump, as well as the fact that undecided voters broke in favor of Trump on Election Day, highlighting the imperfections of polling and the need to present election predictions in a more accurate and nuanced way.
43:05
Polls are not meant to predict who will win an election, but rather to provide a snapshot of how people are feeling at a specific moment and to prevent leaders from deciding what people want or need. Internet polling is a new and disruptive form of polling that is still being figured out, but transparency and understanding the methodology can help determine the reliability of a poll.
47:27
This section does not contain any relevant information about election polling.
Categories:
Society & Culture