Significant disproportionality calculations rely on risk ratios to dive deeper into subgroup differences in outcome data, such as identification, placement, and discipline. Risk ratios are a useful tool in a state’s data analysis toolbox. What if the same tool could be used to identify subgroup differences in other student outcome data? Even better, what if the same tool could be applied to student outcome data that are already collected by states and districts? Let’s explore one potential option for states to examine these subgroup differences using commonly collected data: graduation and dropout.
Examining Differences: Using “Likelihood Ratios” to Understand Graduation and Dropout Patterns
“Likelihood ratios” are mathematically identical to risk ratios used for significant disproportionality calculations. We reframe “risk” into “likelihood” in this case, however, because graduation is a positive outcome and dropout is a negative outcome. While it makes a certain amount of sense to talk about the risk of a negative outcome like dropout, it makes less sense to talk about the “risk” of a positive outcome like graduation (i.e., the “risk” of graduating). Using “likelihood,” we can talk about both outcomes – the likelihood of graduating and the likelihood of dropping out.
Likelihood ratios can tell us the relative likelihood of one subgroup graduating or dropping out as compared to all other students. These ratios allow us to determine how much more likely or how much less likely it is for a subgroup to graduate or dropout as compared to all other students.
Using likelihood ratios, let’s presume that a state found out that the English learner subgroup was twice as likely as all other students to dropout. What’s next? How can states and districts use this information to improve services for English learner students with disabilities?
Promoting Systemic Change: Using Graduation and Dropout Patterns to Inform Transition Planning
One way states and districts can use this information is to consider strategies to improve transition planning, which may, in turn, decrease the likelihood that English learners with disabilities will dropout.
High-quality transition planning → Decreased likelihood of dropout
When disparities in the likelihood of dropout are found, states and districts can drill down into transition plans for a specific subgroup of students. In this example, states could provide guidance to districts on how to drill down into transition plans for English learners. For example, districts could hold student focus groups or partner with parent groups to conduct a root cause analysis. States and districts could also host data conferences or provide pre-data collection technical assistance for districts to examine their transition planning process.
Graduation and dropout data can be used in a feedback loop to inform transition planning and improve outcomes for students with disabilities. Using likelihood ratios gives states another perspective on the differences in outcomes for subgroups of students, which may help districts and schools provide more tailored services and supports for specific subgroups of students with disabilities.
To learn more, watch Session 4 from IDC’s Significant Disproportionality Summit: Beyond Significant Disproportionality.
- Erin Lomax