1. Building an Early Warning System
- When you’ve identified at-risk indicators for students through a predictive analytics study, can you quickly plug those indicators into an Early Warning System that tracks at-risk students over time?
With the right technology solution, once a predictive study has been completed it should be possible to connect the underlying data for those indicators, and build the associated dashboards, in just a few weeks.
- Does your Early Warning System make it easy to see broad, district-level patterns and trends, as well as providing more specific, actionable information for building-level staff? In particular, can teachers and principals quickly identify at-risk students and the specific indicators (such as absences, behavioral incidents, course failure) that identify that student as at risk?
- Can you analyze student data longitudinally, seeing data for the current year as well as trends and patterns in data across multiple years? Can you go back in time and see the courses, assessments, attendance history, and other data for a specific student from prior years?
- Are you able to refresh the data in your Early Warning System on a daily basis?
While data such as course credits changes infrequently, other data such as attendance can change every day. You therefore need an import mechanism that ensures you always have the most up-to-date view of your student data, helping you identify students as soon as they become at-risk.
- Is it possible to define and analyze customized student cohorts such as all 9th grade male students with athletic eligibility or 6th grade Special Ed students with a GPA below 2.0?
Being able to create and analyze such cohorts helps you rapidly answer specific questions that might otherwise take hours to discover using spreadsheets.
2. Predicting Future Performance
- Can you predict which High School student will take more than four years to graduate? Based on this prediction, can you quickly identify the specific students who are not on track, making it easier for school counselors to meet with those students and discuss remediations?
- If you have performed a predictive validity study to determine a relationship between your benchmark and state summative tests, can you use the results from that study to build interactive performance bands for your benchmark assessments?
These results could help you see students who are above or below cut scores, and identify specific students whose results could be improved through targeted interventions.
- Do you know which students are meeting college readiness benchmarks, based on ACT® or SAT® scores? Having identified students who are not meeting benchmark scores, can you identify common patterns in their school history or specific standards where they have struggled?
3. Presenting Information Intelligently
- Is information presented in a way that is very visual and easy to understand?
Visual presentation is critical for usability, and also helps to convey summary information at a glance.
- Is it easy to access detailed information, ideally in the same view as the summarized charts?
While summary views are useful, it’s also important to get down to the details on specific students (for example, seeing their absences, grades, and assessment scores).
- Are charts interactive? For example, can you click on a segment of a pie chart or a section of a performance band to select that specific cohort of students for further analysis?
Interactivity makes it easy to navigate through the data and discover new associations or patterns.
- Can you search your data for specific values? For example, can you find a specific student name or student ID?
Ideally, the search results are immediately filtered as you type, bringing relevant values to the top.
- Can you not only filter your data by a single criterion, but apply multiple filters at once? How easily can you apply a filter showing only 9th Grade Hispanic female students from a particular High School who qualify for Free or Reduced Lunch?
- Is it possible to view information along different dimensions with a single click? For example, can you change a chart to show results by grade level to results by ethnicity to results by school name, each with just a single click?
4. Storing and Accessing Your Data
- Do you need a data warehouse to centralize and store your district-wide data?
Some analytics solutions require a data warehouse to consolidate all of your data into one place. However, building a data warehouse can be time-consuming and expensive, and can require considerable technical expertise. Modern predictive analytics solutions can bypass the need for a separate data warehouse, storing all data in memory where it is fast to access.
- Can you import data from a large number of disparate data sources?
Predictive studies often analyze a large number of variables, identifying those that have the most significant impact on student outcomes. Make sure that your Predictive Analytics solution is able to support all of those data sources.
- Can you export data to a commonly used tool such as Excel with a single click?
It can sometimes be useful to perform additional analysis in an external tool, so make sure you can get the data out without needing to call someone in IT to help you.
Questions to Ask Your Analytics Partner
How much time does it take to answer questions?
- Can you easily see all the data you need when you need it?
- Can you rapidly assemble visualizations for stakeholders to show school or district progress on student achievement? On specific programs?
- Can you quickly identify achievement gaps for specific demographics?
- Can you easily disaggregate your data to answer questions about progress within specific demographics to help close achievement gaps?
- Are you confident that you’ve included all the relevant data in your analysis?
- Can you answer follow-up questions to the original concern without having to generate new views?
- Can you quickly explore associations you see in your data without having to create new charts or spreadsheets?
- Are you dependent on one person for creating all reports and doing all data analysis?
- Can you collaborate rapidly with others on your team to discover connections in your data and gain new insights?
- Is your data presented visually?
- Are you confident that you can access basic student performance data?
- Can you rapidly uncover actionable data without having to pore through spreadsheets or create data warehouse reports?
- Can you rapidly drill down through your data to different levels such as district, school, or student?
“Tying in…all our student data has been instrumental in driving instruction and intervention. Scantron Analytics makes data easy to use—we’re better prepared for ESSA. We’re really feeling ahead of the game.”
Julie Oziemkowski, Director of School Improvement, Community Unit School District 200, Illinois
Can you explore data from multiple sources in the same place at the same time?
- Is your source data in multiple spreadsheets, stored in multiple locations?
- Do you have to manually connect data using spreadsheets?
- Do you have to reach out to data experts to answer questions or can you answer them yourself?
- Do you have to wait for district IT staff to refresh data in a data warehouse?
- Can you easily filter or sort your data? Do you have to manually combine spreadsheets to do so?
Can you easily support decisions using real-time data?
- Can you rapidly incorporate the most recent assessment results, attendance data, discipline information, and student details into accountability reports and material you use to adjust instruction?
- Are you confident that your evidence-based decision-making includes the most recent data?
- Can you identify key achievement patterns in time to take action?
- Is your system accessible—always available to you with minimal effort on your part?
- Can you quickly assess the effectiveness of school turnaround or student achievement initiatives?
- Are you confident that your data supports collaboration with colleagues, leaders, and stakeholders?
- Can you provide the data needed to support just-in-time interventions and track their success?