I recently read Beyond the Horse Race written by my esteemed colleague, pollster, and author John Zogby. Here are some of my personal, key takeaways:
- We should poll as late in the game as possible as more and more voters make up their minds late. We have to hear the grass grow.
- Always expect imperfections because we poll humans. (And, I would like to add, should continue to do so even with AI)
- A survey can tell you much more than who’s ahead. In my own work, the vote question is actually often just a small part of a survey. I test messages, potential, identify target audiences, analyze sources of information, and explore blind biographies.
- After every election, narratives are created in the media about why somebody won or lost. These urban legends are often not grounded in data.
- Journalists and politicians love to hate polling. But they jump on the data.