Intersection

Safety in numbers

Or, how stats are used to prove almost anything

Damon Hodge

Education and crime are driven by numbers. Nothing especially revelatory there. High graduation rates—good. High dropout rates—bad. High crime rates—bad. High arrest rates—good.

So what’s it mean if a district has high dropout rates (around 6 percent for Nevada in 2005, second worst in the nation) and poor graduation stats (consistently a few points below the national average)? And should we cheer a police department for making more arrests in areas generally known for crime?

Let’s take a look.

 Education

The Council for a Better Nevada has been quiet since it failed to muscle New York education standout Eric Nadelstern into the superintendent’s job early last year. But the lower profile hasn’t meant lessened interest in its cause du jour: a wholesale revamp of public schools. Give Council spokeswoman Maureen Peckman an ear on the subject and she’ll take a mile.

Now, as then, two of her biggest complaints center on allegations of grade inflation and padded graduation rates. Something about the district’s slowly improving graduation numbers doesn’t square with her. Too many students need remediation once they’re in college.

Peckman says instructors are under pressure to teach to the test—rather than focus on aptitude—and encouraged to pass along academically marginal students. She says it’s well known that the school district deletes dropouts from its records if they can’t contact them, thus padding graduation numbers.

Not surprisingly, Sue Daellenbach, academic manager with the district’s Assessment and Accountability Division, disagrees. She says there’s no way to fudge graduation numbers: “Students have to have 22 1⁄2 credits and have to pass proficiency in math, reading and science, so there’s no way to get around that. Dropouts are figured into graduation rates.”

Grade inflation, well, that’s tougher to get a handle on. Says Daellenbach: “Our students’ SAT and ACT scores seem consistent with the grades students are getting in classes.

“You have lots of human beings involved in the process of education,” she adds. “We are moving to end-of-course tests in math to test what’s being taught in the classroom.”

No resolution here.

Crime

All’s calm now on Morgan Street. The few remaining families outside Marble Manor are shuffling inside, just in case the foreboding clouds on this overcast Wednesday afternoon decide to open up. Before heading in, a young girl gets in one last round of hopscotch, bouncing in and out of hand-drawn squares on the sidewalk near where, just weeks ago, four kids were shot at a birthday party.

The recent shooting aside, crime numbers seem to indicate an improving climate. Cops credit the Safe Village anti-gang initiative.

Yes, says a man outside Marble Manor’s security office, things are better. But that’s because they weren’t super bad to begin with. Over there—he points to the Treeline Apartments across the street—“ain’t shit changed. They still got ghetto birds [cop helicopters] flying every night looking for somebody.”

Sgt. William Seifert of the Bolden Area Command says a comparison of violent gang crime from February-April 2006 to the same period this year shows fewer attempted murders (nine compared to five); gang-related homicides (two compared to none); knife attacks resulting in injuries (two compared to one); and knife attacks resulting in no injuries (eight compared to five). The only increase was in gang-related shootings: two victims last year (five this year).

So, he says, Safe Village seems to be working.

Or is it?

A search using the CrimeView crime-mapping program on Metro’s website shows the area is still a criminal hotbed: burglary, narcotics, assaults.

Since Safe Village only focuses on violent gang crime, Seifert says “it would not include data on things like domestic violence. We can’t extrapolate this for the entire command.” Plus, the comparison was only for three months. Summer, when temperatures and tempers rise, will provide a sterner test.

So no resolution here.

Conclusion

Numbers can be used to fit a predetermined hypothesis. Nothing earth-shattering there. (See: Enron.) So perhaps the resolution is twofold: The numbers that stories tell often depend on the interests they represent.

  • Get More Stories from Tue, May 8, 2007
Top of Story