How to count crowds accurately, not for PR

Counting large crowds is an inexact science:

“In reality, estimating the size of crowds at mass public events is much more about public relations than a quest for truth,” said Steve Doig, a crowd counting expert who is the Knight Chair in Journalism at Arizona State University.

So how can this be done well?

1. Make a grid
A credible estimate will require knowing the size of the area where the crowd is gathered.

2. Estimate density

It’s important to understand that crowds are not uniform in nature. People clump in some areas and spread out in others. Determining density helps understand how many people can realistically fit into a space…

3. Verify with other sources

A large crowd will require special accommodations. Many will choose to take public transportation to an event. Others will drive. Either way, attempt to compare the crowd-size estimate with other sources, like passenger volume data.

It is not unusual to have vested interests when acquiring data. Different sources with different vantage points – like organizers, police or officials, and the media – could produce multiple counts for a single large event. Perhaps the people with the better social position are the ones whose numbers end up carrying the day. Yet, we could have a variety of reasons for wanting to have the most accurate data including for history’s sake and in order to provide the needed local services for such large gatherings.

Just for fun, here is Wikipedia’s List of largest peaceful gatherings in history. Interestingly, there is a section at the bottom that discusses the methodology of accurate crowd counts. However, it looks like the citations for most of these large crowd counts refer to media sources which could be drawing from a variety of counters including the media itself.

Vested interests in the telling of history

In a recent class, we were discussing an article that talked about a number of issues of doing research after an event has happened. One of the main concerns was the vested interests of the respondents. Particularly, if the respondent were socially near to the event or phenomenon, their retelling is filtered through their own personal interests.

A classic example of this came across the news wires today: some new information that sheds light on why the Titantic hit an iceberg and then sank so quickly. A brief part of the story:

The Titanic hit an iceberg in 1912 because of a basic steering error, and only sank as fast as it did because an official persuaded the captain to continue sailing, an author said in an interview published on Wednesday.

Louise Patten, a writer and granddaughter of Titanic second officer Charles Lightoller, said the truth about what happened nearly 100 years ago had been hidden for fear of tarnishing the reputation of her grandfather, who later became a war hero.

Lightoller, the most senior officer to have survived the disaster, covered up the error in two inquiries on both sides of the Atlantic because he was worried it would bankrupt the ill-fated liner’s owners and put his colleagues out of a job.

I’ll be curious to see how quickly this new information can (or cannot) be corroborated.