Louisiana School District Uses Data Mining To Analyze Troublemakers
A year's worth of disciplinary reports are scoured for patterns in behavior and punishment.
In Lafourche Parish, La., middle schools supervisor Chris Bowman wants to understand how students get into trouble and what happens when they do. Bowman had a brainstorm: He took the 33,000 disciplinary reports filed last year by teachers and administrators and ran them through the school district's data mining system.
It occurred to Bowman that analyzing the reports could provide insight into the root causes of disciplinary problems--tardiness, dress code violations, fighting, vandalism, and more--as well as the effectiveness of how the schools deal with them, and how fairly discipline is enforced. The reports generally contain structured data, including demographic information such as the student's grade and age, and a written narrative that describes the infraction.
"I knew there was a wealth of data in these reports," Bowman says. "Inside all that information, in the narratives, were things I wanted to know about." The narratives include detailed descriptions of incidents, how teachers and administrators responded, and disciplinary actions. "I wanted to stop looking at anecdotal evidence alone and start looking at actual data patterns," he says.
The school system, which has 15,000 students in 30 schools, had been using SPSS' data mining software for several years to analyze student test scores. Last year, administrators used SPSS' Text Analysis For Surveys to assess the results of a survey of 450 teachers, administrators, and other school employees.
(click image for larger view)
The key to better behavior may be in better data analysis.
That experience gave Bowman the idea of applying the same technology to analyze the discipline reports. Bowman is using text analysis to search for and categorize key words and see if patterns emerge. A search for the word "uniform," for example, might show if teachers or schools are being overly strict in enforcing dress codes, while a search on "tardy" could identify habitually late students. Bowman can search the records of a student, school, or the entire district.
Bowman combines those findings with demographic data to look for additional patterns. "I'm trying to look beyond the obvious and find out if there are hidden factors in why things happen," he says. Are kids from low-income homes, for example, disciplined for dress code violations?
One thing Bowman hopes to understand is how seemingly minor incidents such as dress code violations or use of foul language escalate into more serious incidents. He hopes to shed light on the effectiveness of "intervention"--punishment such as detention or suspension--based on how often a student gets into trouble despite disciplinary actions.
The Lafourche Parish schools' use of text analysis and data mining in this way is unique, but a number of school systems around the country use data mining tools to study student test results. SPSS is working with research firms Analytic Focus and Reveal Technologies to develop models for predicting student performance in grades K-12 in six school districts in Alabama, Colorado, Iowa, Minnesota, and New York. Meeting the stiff requirements of the No Child Left Behind Act is a major driver.
Bowman began experimenting with the SPSS software in the middle of the last academic year, when disciplinary reports hit 17,000, and is researching all reports filed during the year. He's refining the analysis, and results so far are preliminary, but there are indications that not all students are treated equally. Bowman expects that the findings, which he plans to put into a report for school administrators, will anger some people in the school system.
Building A Mobile Business MindsetAmong 688 respondents, 46% have deployed mobile apps, with an additional 24% planning to in the next year. Soon all apps will look like mobile apps – and it's past time for those with no plans to get cracking.