Our alumnus Shubhabrata (Joy) Mukherjee, PhD, a research assis
A team of researchers from the University of Louisiana at Lafayette who attended the city’s recent Mardi Gras King’s Parade wasn’t interested in catching beads or doubloons.
It was after data.
The group of four, which included students and postdoctoral researchers, fanned out along the parade route. The floats rolled from near the intersection of Jefferson, Surrey and Simcoe streets, and ended near Cajun Field. The researchers walked along the entire 4-mile route.
Their assignment was to estimate crowd size at the request of the Lafayette Police Department. The project will assist law enforcement with crowd management, officer allocation, and emergency response plans for future parades and other events.
Marina Ledet, a senior from Abbeville, La., majoring in mathematics, didn’t mind the assignment. “I actually don’t like parades, but I enjoyed this one because I got to do math,” she said.
Before the researchers hit the parade route, they used software to generate a sampling scheme. It included a predetermined number of steps they would take before stopping to count parade attendees.
At each 100-step interval, they counted spectators standing behind corresponding barricades lining both sides of the parade route. They recorded the numbers in small notebooks, except for sections of large numbers of people. For those, they snapped photos and viewed them later to tally spectators.
After determining a mean number of people behind each barricade, that average was multiplied by the total number of barricades.
Ngan Hoang Nguyen Thuy, a graduate student from Vietnam pursuing her doctoral degree in math, said compiling figures at a noisy, festive parade was easy. Explaining to some spectators why their photos were being taken was harder.
“Families with small children often felt a little bit uncomfortable, until I explained the purpose of the study, which will benefit the community,” she said.
Suntaree Unhapipat, a postdoctoral statistics researcher from Thailand, said team members used two statistical models to find possible ranges for the number of spectators on each side of the route. “We came up with almost identical results each time,” she said.
The researchers estimated the total crowd size for the event at 47,234 people.
That averages out to nearly 12,000 people per mile, or nearly 1,000 per standard city block.
The team, which also included Phontita Thiuthad, a postdoctoral statistics researcher from Thailand, presented the results to LPD representatives, said Dr. Nabendu Pal, a professor of statistics who led the project.
“Worldwide, crowd size is becoming an increasing public safety concern,” Pal explained, referring to recent terrorist attacks in Nice, Berlin, and London. “There are other implications as well. Knowing how many people attended a large gathering compared to how many, for example, needed medical attention can help paramedics marshal resources.”
LPD Corporal Karl Ratcliff said law enforcement efforts to collect similar data can be hindered by officers’ duties, or an incident that diverts their focus.
“We can do the estimates ourselves, but we are often consumed with ensuring safety. It’s good to have eyes from the outside and to be able to put some science behind it,” Ratcliff said.
“It’s a great service to the community, especially at this time of year when we experience large crowds and attract people from around the world at events like Mardi Gras and festivals. We want to do everything we can to be proactive.”