Large public databases and data mining techniques make possible measurements that were once impossible — or too expensive — to pursue. Researchers use administrative data collected for other purposes, such as student test scores, criminal arrest records and healthcare expenditures, to conduct randomized control trials and establish what works. It becomes possible to understand which early interventions make the most difference, and which mix of services under what circumstances help individual clients.
Now let’s hear from you. Have an example? A story? A comment?
Agencies such as the US Administration on Children and Families encourage agencies to shift resources from interventions proven to have little impact—generic parenting classes or counseling, for instance—to those that genuinely improve behavior and emotional health.