How big data is battling child abuse
Technology and data unite every aspect of our lives, and the digital information that we produce is being put to new uses every day. Predictive analytics are used by major card companies to detect fraud, while marketers use data to make campaigns more effective and determine who their target audiences are. With everyone leaving such distinct digital trails, analysis of these digital paper trails is being put to use to improve the lives of children.
Big data is starting to be put to use in support of social workers’ goals of providing help and support to those in need, and in volatile situations. Analytics uses collected and new data to identify patterns, predict behaviors, and help focus the resources of the child welfare system where they can do the most good. But how are they doing it?
Organizations across the world use predictive analytics, to the point that it is vital to the success of any business. Predictive analytics takes a mass of information from multiple sources and runs it through computer systems to sort, analyze, and process. Then, when new information is added to the system, we are able to compare it to previous information.
These systems use what we already know about similar data, and project the most likely path for new data. Credit card companies use this method to detect fraud, and big data is now being used to amass information regarding child behavior, living situation, reports, and investigations to predict the risk of future and recurring abuse.
Approximately seven million children are brought to the attention of child welfare services annually, and new reports are constantly circulated in the media about investigations into the deaths of children due to abuse. The use of predictive analytics is being applied to help aid call screeners in determining if a case file should be opened. Previously this decision was left up to the judgment of call screeners, but the new process can help screeners and their superiors make more informed decisions.
After receiving a phone call, screeners take the necessary details and input them and the child’s identity into their system. The program then determines a risk score between one and 20. The closer to 20 the child’s score is, the more likely they are to be in an unsafe situation. The analytics consider the parents’ prior offenses, reports that have been made, and many other points of related information about the child and their relationship to the people in question.
Ultimately the decision of whether or not to open a case is made by screeners in collaboration with supervisors. Los Angeles is already implementing a version of this system to help make more informed decisions, and while it is not a silver bullet to fix an admittedly messy system, it is helping to focus resources. These benefits are not without their potential downfalls, however.
As is always the case with the uses of Big Data, there are concerns about how these systems may be implicitly biased, and the information misused. Without building in context, disenfranchised families and communities that already struggle with the legal system could be left open to unwarranted scrutiny while leaving higher-income, white perpetrators unidentified.
