Multi-stakeholder analytics partnership using improvement science with strict data governance¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The report recommends that districts, technology providers, and researchers form partnerships that use learning analytics for continuous improvement rather than evaluation. All data is de-identified before entering a secure Azure SQL database within the district's own cloud subscription, and analyses apply controls for demographics, attendance, prior scores, and school climate. The approach follows "an 'improvement science' approach, rather than seeing the work as evaluation or research," using analytics to determine what works and feed that back into the intervention.

Design Implications¶

Context¶

Requirements¶

• a legal agreement covering data management and intellectual property; no personally identifiable information shared between partners; trust built over time among partners

Constraints¶

• the technical, data security and privacy compliance elements of such a partnership are highly complex

Target Learners¶

• school and district leaders

Target Learning Goals¶

• using learning analytics to improve instructional initiatives

Affordances¶

• Seven Elements Effective Pd Framework

Related Strategies¶

• After a study, present final data in school-useful forms, remain available for ongoing conversations, and plan carefully to avoid data overload
• During a study, researchers should visit regularly, give back to the school community, minimize classroom disruption, and share initial data for feedback
• Deliberately bridge student experience data work across tools, initiatives, and partner organizations
• Form design teams by building trust and relationships and prioritizing meaningful representation before project kick-off.
• Districts should share sufficient subgroup data with assessment providers to enable multi-subgroup disaggregation

Examples¶

-

Key Sources¶

• Peters, V., Means, B., Langworthy, M., Neufeld, P., Coe, R., Meehan, K., & Smith, S. (2018). Enabling Analytics for Improvement: Lessons from Year 2 of Fresno's Personalized Learning Initiative. Digital Promise. https://digitalpromise.dspacedirect.org/items/f5ad25ac-1bc9-4a05-9a4e-24c25ab19e8d. doi:10.51388/20.500.12265/53