A new study has developed a novel way of evaluating and improving VAMs. By taking data from Boston schools with admissions lotteries, the scholars have used the random assignment of students to schools to see how similar groups of students fare in different classroom settings.