The Pennsylvania State University delegation — a professor, three graduate students, and 10 undergrads — arrived here at the Republican National Convention full of excitement suffused with dread.
Political junkies of all stripes were descending on Cleveland to reckon with the once-unthinkable coronation of the businessman Donald J. Trump as the standard-bearer of the Republican Party. The candidate’s showmanship, combined with rumblings of a possible mutiny among the delegates, promised a spectacle inside the heavily guarded Quicken Loans Arena.
But the Penn State group was not here for the party. They were here for the party-crashers.
The team, led by Lee Ann Banaszak, a professor of political science, came to collect data for a study of the protesters — not just what they want and when they want it but also: Where are they from? How long are they staying? What do they care about?
For all the public fascination with protesters, reliable answers to those basic questions are actually rare, said Ms. Banaszak. Other researchers have tried handing out surveys at political conventions to ask about protest activity, she said, but that approach does not catch a lot of good data about the people who actually show up for rallies.
And so the professor came here to lead a crew of wide-eyed undergrads into the maw of democracy in search of a better understanding of a fundamental political pastime.
With luck, they would make it out in one piece.
The hype around the RNC protests began in March, when activists mobbed a Trump rally in Chicago and forced it to shut down. In June there were outbreaks of violence outside a Trump rally in San Jose, Calif. The chattering classes murmured about a possible repeat of 1968, when protesters clashed with the police outside the Democratic National Convention, in Chicago. As the Republican convention drew near, Cleveland called in reinforcements from around the country and braced for riots and mass arrests.
The Penn State researchers, too, had prepared for the worst. Most of them had never been to a protest. They were not old enough to remember the genteel friction of Bush v. Gore, let alone the chaos of ’68. They arrived in town armed with surveys and shirts the color of lemon-lime Gatorade with "RESEARCHER" printed conspicuously across their backs.
On Sunday night, the eve of the convention, the students stood on a patio outside the Alpha Phi sorority house at Case Western Reserve University, where they were staying, and nervously rehearsed their scripts.
The multiple-choice survey they were supposed to administer was brief but wordy. It was all students could do to get the words right, let alone approximate a natural delivery. Aviva Doery, a rising senior with short brown hair, accused her reading partner of cracking her up. "You’re always smiling at me!" she said.
Distractions of the field, of course, would be more intense. The heavy police presence had been obvious from the moment the team arrived at Case Western. The university was housing nearly 2,000 police officers who had traveled from out of town in anticipation of protest-related tumult; they roved the University Circle neighborhood, smiling and cracking jokes like packs of giddy freshmen. Ms. Doery and her teammates had checked in at the university next to what looked like a SWAT team.
The danger of bringing student-researchers into the circus atmosphere of the RNC was not lost on Ms. Banaszak. Driving into town, she had spotted a line of officers marching across an empty football field in formation, batons drawn. They were practicing, too.
Inside the sorority house, the professor and her lieutenants, a trio of graduate students, quizzed the team on safety.
What are you supposed to do if you see police officers approaching a protest in formation?
"Leave," answered the students in unison.
What if you see a protester hit another protester? Leave. Feel unsafe for any reason? Leave. See someone get arrested? Leave.
And if you’re the one being arrested? … Stay. Don’t flee, don’t resist.
The team leaders had taken measures to avoid such a scenario. The message of the students’ "RESEARCHER" T-shirts was, essentially, "don’t arrest us." But Kevin Reuning, a graduate student, had taken things a step further: He had fashioned official-looking badges that even included the study’s institutional review board code, a reference so obscure that even the students didn’t know what it meant. "I needed it to look authentic," said Mr. Reuning, laughing.
Before the students retired for the night, the team leaders reminded them to pack rain gear. The forecast called for storms.
"Anyone’s parents freaking out at all about this?" asked Mr. Reuning.
Some convention protesters do whatever it takes to stand out in a crowd.
They are the dueling street preachers; the climate-change activist in the polar-bear costume; the fratty-looking guy in a cowboy hat walking with a leash around the neck of another fratty-looking guy, who is on all fours and wearing a sign that says "Kick my dog 4 Trump." ("A lot of people don’t understand our message," Cowboy Guy drawled to amused onlookers. "The message is chaos.")
The goal is to bait one of the thousands of reporters — or, really, anyone with an iPhone and a Twitter account — into gawking. In a staging ground for grand displays of political loyalty, the real bias is for the weird.
Ms. Banaszak’s goal is to cut through that bias. Her quarry is the statistical whole, not the sensational outliers. If chaos is the dominant ethos around the fringes of the convention, the Penn State team might have the most radical agenda of all: dispassionate census-taking.
This is how they do it: When the Penn State team rolls up on a protest, the professor and her lieutenants eyeball the crowd and make a rough estimate of its size and shape. If the crowd is large enough to get a good sample, they’ll dispatch the students to interview people. But not just anyone. Using her best sense of the protest’s geometry, Ms. Banaszak will count protesters at consistent intervals, and then will send a student to interview each randomly selected protester.
A march, for example, could be seen in rows. Ms. Banaszak might count five people across, two people down, then send a student to interview that protester. The professor would then count another five across, two down, and dispatch another student. And so on.
In the field, of course, things could get messy. People don’t often stand in perfect rows. They clump. The researchers would have to adjust and do their best to maintain the integrity of the system.
"We’re going to get a representation of what’s outside," said Mr. Reuning. "It’s not going to be perfect. Social-science research never is."
The next day, megaphone distortion emanated from a fountain on a mall in downtown Cleveland where a "Stop Trump!" rally had been scheduled for midday. The Penn State team, a slice of lemon-lime in a sea of red, white, and blue, moved toward the din.
"That sounds protest-y!" said Ms. Doery hopefully.
Ms. Banaszak eyeballed the crowd. Ideally she would want a group of at least 300, to increase the likelihood of getting a good data sample. This crowd was a little thin. "We should still try," she told one of her grad students.
They decided the best method would be to start at the corners and work their way in. Ms. Doery gripped her iPhone, which had a data-collection app loaded with the survey. "Do we want to practice real quick?" she asked a teammate.
No, it was game time.
"You’re on the woman with the white jacket," said the professor, pointing through the crowd.
Ms. Doery nodded, marched up to the woman, and introduced herself as a Penn State researcher.
Carefully but quickly, she recited a version of the script. Throughout the day, the message barely changed: We’re out here surveying protesters, and we would like to know a little more about the matters close to you. So we would like to do a two- to three-minute survey. It’s completely voluntary, and your answers will be kept completely confidential. Can I ask you a few questions?
White Jacket said she couldn’t. She was with the press.
"But we’re not!" chimed in her neighbor, an eager young woman wearing a red T-shirt with a pro-labor theme.
Ms. Doery didn’t miss a beat. "Sorry, it’s randomly selected," she said. "I have to go back to my team leader."
Before long she got a hit: an older woman in a white running shirt and sunglasses, holding a "Stop Trump" sign.
Why was she here? To express solidarity. What was her biggest issue? Trump as the nominee. How many hours was she planning to stick around the convention? She doesn’t know, a couple. Where’s she from? West Cleveland. Would she be willing to give her email for a possible follow-up survey? Yes, OK.
Running Shirt’s answers went into the app and then into the cloud, where Ms. Banaszak could retrieve them later.
"We got the power!" chanted protesters nearby.
After getting a "No" from a guy who looked like an undercover cop and a "Yes" from a woman wearing a faux-floral headband with American-flag petals, Ms. Doery thought she’d hit the jackpot — a man with a tall hat waving a blow-up doll with a Trump hairpiece. The man was holding a sign: "Small Hands, Small Mind, Big A-Hole."
"Yesss," she said as Ms. Banaszak pointed him out. "I was hoping it would be that guy."
Alas, Blow-Up Doll Guy turned out to be an underwhelming interview. Ms. Doery read him a menu of possible answers to the question of why he had come out to protest, but he was wiggling his doll for the cameras and lost track of the options.
"What was the first one?" he asked.
"Pressuring politicians to make changes?"
"Yeah," said Blow-Up Doll Guy, "the whole damn thing needs changing!"
After about a half-hour of surveying, the team reconvened next to a fountain. The professor and her lieutenant traded notes on how they had tried to get random samples from the relatively spread-out, shapeless crowd. Ms. Doery and her teammates swapped stories.
Most people in the crowd were agreeable, but not everyone. "I don’t want any part of this college stuff," someone had told one member of the team.
He meant the research, not the protest.
Nineteen sixty-eight this was not. Well-attended protests turned out to be elusive.
After the anti-Trump rally, the Penn State team went to a pro-Trump rally, which was located — providentially, for everyone except the researchers — on the other side of town. The rally had a permit for 5,000 attendees, but only a few hundred people were lounging among "Hillary for Prison" signs on a green slope overlooking the Cuyahoga River. Dozens appeared to be reporters.
That afternoon Ms. Banaszak led her team on a sweltering hike across the river in search of a rally for Iraq War veterans, only to find nothing but a trash-can percussion band on wheels ("We do not support Trump, but we do support his haircut!") and a bunch of bored police officers shooting the breeze.
Dead end, said Ms. Banaszak. It happens.
The afternoon was not a total loss: Mr. Reuning texted the professor to report that his group, which had struck out that morning at a series of mostly-empty "free-speech zones," had found a crowd of more than 500 at an End Poverty Now march.
On the way back to town, Ms. Banaszak’s team relaxed a bit. The sun was out; there had been no storms after all. They passed a firebrand preaching a sermon through a megaphone for a crowd too small to survey, and spotted Blow-Up Doll Guy cooling his heels nearby. A plume of smoke rose from what turned out to be a food truck. Some police officers moseyed by on horseback.
Hey, said the professor, remember the rule for what to do if you see the police on horses?
Not long after that, they did.