Getting through the geek talk: Analyst explains St. Paul school discipline data
Alleen Brown, Twin Cities Daily Planet, April 15, 2012 –
Judging from the rhetoric coming from some educators, data is going to save us all. They’re right that it might help, but not if we don’t understand it.
Last Wednesday, the Twin Cities Research Group, a network of researchers and data analysts, met at the Wilder Foundation to hear St. Paul Public Schools research analyst Cindy Porter talk about suspension rates. She pointed out some of the challenges to collecting clear and actionable suspension data, and she highlighted some interesting numbers.
Why do we care about suspensions? Porter said St. Paul has not done a formal analysis connecting high suspension rates to low graduation rates, but the connection is self-evident and widely accepted.
The cause for the connection is not as widely agreed upon, but Porter said one thing is certain: attendance is the single greatest predictor of academic outcome. “If you’re suspended, or you’re even sent down to the office, you’re not in the classroom, and you’re not learning,” she said. “If we do a suspension, it’s an action that we are taking to remove the student from instruction.”
In St. Paul, suspensions have actually gone down in the past few years. In the 2007-2008 school year 20 percent of St. Paul junior high students were suspended. In 2008-2009, the rate dropped to 14 percent. Rates dropped in elementary and high school that year as well. They stayed consistent the next two years. Porter said data newer than 2010 was not ready for presentation.
Other data indicate serious issues. A disparity in suspension rates between white and non-white students has incited alarm locally and nationally. Indeed, the data Porter presented showed that St. Paul African American and American Indian students are more likely to be suspended than white or Asian students.
In the first two quarters of the 2010-2011 school year, 70 percent of suspensions were of African American students, although they only represented 30 percent of enrollment. Only five percent of suspensions were of Asian students, although they represented 30 percent of enrollment, and 11 percent of suspensions were of white students, though they made up 24 percent of enrollment.
Porter acknowledged that the numbers don’t tell the whole story. Are some demographic kid-groups worse-behaved than others? Or do we have racist teachers? The numbers don’t tell us. Do suspension rates vary between schools because one school doesn’t have money for an in-school suspension classroom? The numbers don’t tell us.
The St. Paul district is working hard to eliminate as much variation in adult judgment as possible. At the school level, administrators are implementing PBIS, or Positive Behavioral Interventions and Supports, a model that requires teachers and administrators to use discipline data to evaluate the effectiveness of far-reaching school policies.
On a district level, Porter passed around the district’s discipline handbook, which includes a matrix of behaviors rated on a scale of severity from 1 to 5. Each level is matched with recommended consequences.
The matrix, developed with input from school administrators, only recommends suspension for severity levels 4 and 5. Level 5 behaviors include things like possession of a firearm or aggravated assault, and they require suspension and an expulsion referral. Level 4 behaviors include anything from bullying to drug dealing to disorderly conduct, and suspension is one of many possible interventions.
In practice, adult and kid behaviors fall outside of many of the matrix’s suggestions. Porter displayed data showing that 28 percent of suspensions in 2009-2010 responded to behaviors for which policy does not recommend suspension. Level 3 behaviors include things like physical aggression causing no bodily harm, defiance and disruptive behavior.
Of course, no system could neatly categorize the levels of creative behavior kids come up with. As Porter suggested, physical aggression could mean elbowing or a punch. A behavior categorized as level 2, might be immersed in circumstances that make suspension appropriate.
Adult judgment is a factor that can’t be eliminated from the equation. “What it looks like to me sitting at my computer is not what it looks like to the person who is dealing with this student daily,” Porter said.