As colleges seek ways to make their campuses safer, many have opted to examine the criminal histories of students before they're admitted. New research, however, reveals that criminal-background checks and pre-admission screening do not accurately predict whether an incoming student will pose a threat or disruption in college.
Based on an analysis of nearly 7,000 seniors at a large Southern university, a report says that only 3 percent of students who engaged in misconduct on the campus during their college years had reported criminal histories during the admissions process. Of the students who did report a criminal record, meanwhile, just under 9 percent were accused of misconduct during college.
The report, published in the journal Injury Prevention in February, was written by Carol W. Runyan, an epidemiologist at the Colorado School of Public Health, and three other researchers.
For years, colleges and legal experts have wrestled with the question of whether—and how—institutions should attempt to identify incoming students who might present a threat to public safety. The quandary is what to do with any information collected: how to evaluate it fairly and consistently while avoiding discrimination against some students but also protecting against any future incident.
According to a national survey in 2010, more than 60 percent of colleges consider applicants' criminal histories in admissions decisions, but only half of those colleges have formal policies on how to do so, and only 38 percent of admissions staffs receive training on interpreting criminal records.
The new research sought to examine if students who were likely to engage in misconduct could be effectively screened during the application process. It also explored whether students with a criminal background upon entering college were more likely to commit crimes while enrolled than were students who started with clean records.
Researchers reviewed students' responses to application questions about their criminal history, which asked them to say whether they'd been convicted, taken responsibility for a crime, or had charges pending against them at that time. A "yes" to any of those questions meant the students were considered to have criminal histories.
To evaluate students' behavior in college, the researchers looked at the university's disciplinary records and kept track of nonacademic misconduct violations, focusing on offenses like assault, robbery, property crimes, driving under the influence, marijuana use, and other drug-related charges. They also included cases that the institution's honor court had dismissed but that were prosecuted successfully in local court. (The report states that the research was approved by the institutional review board at the University of North Carolina.)
The findings reveal that students who were guilty of misconduct in college were more likely than their classmates to have had pre-college criminal records. But the screening questions often did not identify which students would go on to commit crimes, and most students who did have records before enrolling in college didn't cause any trouble once there.
In the report, Ms. Runyan points out that the research "raises as many questions as it answers." Many questions, she says, are practical and ethical: If colleges are going to make smart decisions about pre-admission screening, she writes, they'll need to think about how past behavior influences future actions. And even if the screening does accurately identify "likely troublemakers," colleges must decide in which cases to admit them.
'What You're Seeing'
Some institutions, including the University of North Carolina system and the University of Virginia, have considered investigating students' criminal backgrounds in the wake of a tragedy. But the practice of admissions screening and background checks aren't effective tools in predicting future misconduct, says Peter F. Lake, director of the Center for Excellence in Higher Education Law and Policy at Stetson University.
"You're really only going to net a relatively small number of people," Mr. Lake says. "When you do get stuff, a lot of it requires a really careful understanding of the criminal-justice system to know what you're seeing."
In some cases, he says, the criminal actions that show up on a person's record indicate no violent behavior. A series of convictions for minor drug-possession offenses, for instance, will appear on a prior record, but dropped charges for domestic violence won't. The latter, Mr. Lake says, doesn't technically constitute a criminal record—but it would be far more troubling to him, and perhaps more predictive of future violence.
For more-constructive clues, he says, colleges should consider examining students' grade-school, high-school, and military records. "When you dig into those records, you find a whole different set of stories that are much more likely to give you a perspective on someone who might be accelerating toward more-dangerous behavior."
The new research leaves unanswered many lingering questions about such admissions screening, says Ada Meloy, general counsel at the American Council on Education.
Colleges don't know how best to process the information they collect, Ms. Meloy says. It's reasonable to ask a prospective student about his past, she says. "But in asking the question, an institution should have in mind what it's going to do with the answer."
Given the complexity of the criminal-justice system—and the range of notations that appear (or not) on a person's record—reacting to information can be a challenge, she says.
And what colleges want to avoid, both legal experts agree, is using background checks and screening questions to shut out students for whom college could be a positive experience.
"We see people who transform," Mr. Lake says. "They mess up in high school, get caught with pot or get in a fight in a bar, and they become upstanding citizens, in some cases very quickly."