Research for media reporting on RCV elections

In the first part of our research, we tested a variety of ways press articles display the results RCV elections using 10 samples. In the second part, we narrowed the sample displays we tested down to two.

Here are a few of our key findings:

Most participants described their favorite samples as “simple.” 

• By simple, they meant that the results were quick to scan through.
• Even in less popular samples, checkmarks were cited as simple elements within a complex layout.  
• Most participants cited samples with many data sets as unnecessary, overly complex, and for academics.

Maps give context to understand election results.

• Most participants mentioned that a map helps them understand data 

“I didn’t choose this in my top [choices] because there was nothing I could associate with. Some pictures. Like who’s who, or who’s on top, or a picture of Alaska”

Research participant

For election results displays, people look at visuals first, then at the header, then text.

The information about the election should include:

• Office
• Location
• Status of results

Information about RCV should include:

• A definition of what inactive ballots are and what happens with them
• How many rounds there were by numbering them

This research was conducted by Emma Werowinski and Fernando Sánchez.

What did we want to learn?

• Which best practices help people understand RCV election results?
• Do people understand the status of the results?
• Do people understand who won and how?

What kind of testing did we do?

We conducted intercept tests in NYC.

Where we tested

We tested in neighborhoods that had low-ranking participation in 2021:

• Bushwick, Brooklyn
• Concourse Village, Bronx
• Fordham Heights/Fordham Manor, Bronx