Identify At-Risk Students

During School Year

Before diving into the processes involved in Identify At-Risk Students, we encourage your team to ensure your technical infrastructure is ready for this effort. In the case where you aren’t sure, this checklist will generate a set of questions you can ask the district or provider for clarity. In the case where you know the system(s) cannot do something, this checklist offers considerations for building workaround solutions.

Technical Infrastructure Choose Plays Examples in Practice

Technical Infrastructure

0 of 4 answered

0% complete

Data Integrity & Governance

Is your system set up to automatically update student profiles when new data, including predictive and well-being indicators, become available?

Data Integrity & Governance

Is your system set up to display date of latest update for all data?

Actionable Analysis & Logic

Is your system set up to automatically flag students that cross thresholds for each holistic indicator?

Workaround Option 1:
Check the system FAQs and resources to determine whether/how data can be flagged to show when students are nearing or changing off-track status for each relevant indicator and holistically across indicators. Adjust settings and display options as necessary (and permitted) to enable data-specific flags and triggers.
Workaround Option 2:
If the system will only automate flags for some types of data or restricts users to a limited number of flags, it’s possible to export a subset of the data into Excel and use conditional formatting to visually mark values as above and below one or multiple thresholds. If this data is collected several times, an Excel formula to capture anywhere a student’s status has changed and what type of change it was (e.g., moving back on track vs moving off track) can be helpful. Analytics tools such as Power BI or Tableau are also designed to report out on and display threshold data if the centrallly provided system will not support it. Ideally, set flags in the primary system for data that is higher-impact, more transactional, and when there valuable insights or actions tied to real-time alerts. Data that change infrequently may be lower priority.
Ai-enabled workaround:
If the system cannot flag data based on specific conditions, predictive indicator data and data tied to graduation requirements should be exported on a regular cadence and formatted in Excel for readability (e.g., conditional formatting, charts, relevant calculations and aggregation) or an analytics system should be put into place for ongoing use. Unless a databridge or automated integration is available, the data that should be prioritized for external analysis should be those that are updated at key points in the year. Without real-time, automated integration, most attendance, intervention progress, assigment-level, and other transactional data would be too cumbersome to manage outside of the systems in which they are collected and as such, must be analyzed and understood separately. If the district has a closed and secure AI tool and allows the upload of student identifying information and data, consider using AI to format and analyze data exports from a variety of sources and to add visible flags when specified data is below, within, or above specified thresholds. Ensure responsible use of AI by referring to the Considerations for AI Use section of this playbook.

Questions for Provider:

Can users set custom thresholds based on a combination of survey scores and academic data?
Once a flag is triggered, is there a built-in workflow to “acknowledge” or “resolve” the flag, or does it remain active until the student’s data naturally improve?
Does the platform track how long a student has remained above a certain threshold, and can users pull reports based on that duration?
Is there a log showing exactly when a flag was triggered and which specific data point (e.g., a survey response vs an attendance event) caused the threshold to be crossed?

Questions for District:

Which data have thresholds and which are set up to send automated notifications?
Is there a way to group, batch, or customize notifications to streamline tracking and response for staff?
How can users track the status of a flag within a student’s profile?

Intuitive Insights & Dashboards

Is your system set up to generate dynamic visualizations and dashboards designed to help staff consider cluster or group intervention strategies for students identified as off track?

Workaround Option 1:
Check the system FAQs and resources to determine whether/how data visualizations can be adapted, customized, or added —and by whom. Some commonly found visualizations that support analysis and grouping include: scatterplots (compare values between groups on multiple measures), vertical stacked bar charts (look at the distribution of a trend over time—e.g., % interventions assigned for math, ELA, both, pr neither at specific points in time), or box and whisker plot (overlay data with lines showing min, max, and median values). Adjust or add visualizations to meet user needs.
Workaround Option 2:
For simpler data or visualization needs, export relevant data into Excel and use the chart functionality to customize one or multiple displays. To support multi-variable displays, longitudinal reporting, or more robust drilldown capability, consider a data analytics tool such as Power BI or Tableau.
Ai-enabled workaround:
If the district has a closed and secure AI tool and allows the upload of student identifying information and data, export the raw data needed for student success team tracking and discussion, identify specific visualizations or chart types for the data, and prompt the AI tool to generate the visualizations. Remember to specify how the data should be labeled, how outlier data values should be treated (e.g., if there is a null value, treat it like a zero), and how it should distingush between variables (e.g., colors, symbols). Prompt the AI to explain the steps taken to compile the data and to share any inconsistencies, risks, or issues with the data and how they have been resolved or attended to in the resulting visualization. If any of the explanations are not clear or do not make sense, prompt the AI to alter the approach it takes to developing the visualization. If the visualization is rerun, prompt the AI to call out the specific changes in each iteration from the prior version(s). Remember to at least spot check the data against the source data to ensure accuracy. Ensure responsible use of AI by referring to the Considerations for AI Use section of this playbook.

Questions for Provider:

Can the platform automatically highlight/pull out outliers, such as students whose scores are significantly different from other students?
Are visualizations interactive, allowing users to click into a data point to see more details, such as specific students included, analysis, and more? To what extent can the district customize visualizations?
How does the platform visualize trends to show a student’s improving or delining well-being and academic performance?

Questions for District:

Does the district already have tools available to turn raw well-being and academic data into visual charts or heat maps for easier analysis?
What visualizations are most important for supporting student success discussion?

Continue to Plays