class: center, middle, inverse, title-slide .title[ # Using Data to Identify and Address Housing Instability ] .author[ ### Michael LaClair ] .institute[ ### Guilford County Senior Data Analyst Interview ] --- #The Objective ##Determine whether data can help identify housing instability early enough to connect students and families with effective support. --- # Understand .flex-container[ .flex-box[ ### Frame - Problem - Community - Stakeholders - County and School District Leadership - GCS and County Departments - Community Partners ] .flex-box[ ### Define - Actions - Outcomes ] .flex-box[ ### Decide - Proceed - Redirect - Stop ] .flex-box.different-color[ ### Milestones - Theory of Action - Logic Model ] ] ??? - Frame: In this case we are exploring the context and boundaries of our project. Exploring what is needed and what is possible. Also, we are establishing important relationships. It is important that our partners feel heard. That we are doing this with them, and not to them. Speaking with leadership is important because it can help us identify project champions - Stakeholders - County leadership: goals, capacity, decision criteria - GCS/schools: operational needs and feasible interventions - Community partners: available services/referrals ### Mention GCS SNS example and busdrivers - DEFINE: Once the issues has been framed, we can identify outcomes and the actions and inputs that support those outcomes. - Decide: At this point, do we have stakeholder buy in and a viable plan to address the issue. Perhaps we need to modify our approach/scope/scale (focus more broadly on housing insecurity). What changes do we need to make. - Theory of Action is the hypothesis: "if this is done, this will happen." - The Logic Model is the concrete inputs, actions and outputs. --- #Assess .flex-container[ .flex-box[ ### Govern - Privacy - Security - Data Sharing ] .flex-box[ ### Assess - Data Sources - Availability - Quality - Timeliness - Detail ] .flex-box[ ### Validate - Signal - Bias/Equity - Usefulness ] .flex-box.different-color[ ### Milestones - Data Governance - Data Model - Data Support Deliverables ] ] ??? - At this point, we would do some exploratory analyses and models to test assumptions. --- # Housing Vulnerability in Guilford County (Illustrative Example Only)
??? - The map is just a demonstration of a type of profile that can be created using public datasets. - However, the map also can be used to highlight the importance of determining the meaningfulness of the signal, as well as hidden bias or issues with equity. The map makes a lot of unsupported assumptions: - the data focuses only on renters, what about homeowners. - The weighting of the variables is abitrary (40-20-40) - The cut scores for the vulnerability levels are arbitrary (20-50-80) - The zip codes may not be granular enough (correlation effect size issue) used only because of the data sources available --- #Operationalize .flex-container[ .flex-box[ ### Design - Roles - Responsibilities - Workflow ] .flex-box[ ### Build - Data Processes - Reporting - Technical Requirements ] .flex-box[ ### Act - Interventions - Referrals - Supports ] .flex-box.different-color[ ### Milestones - Operational Workflow - Data & Reporting Processes - Intervention & Referral Processes ] ] --- #Evaluate & Improve .flex-container[ .flex-box[ ### Monitor - Implementation - Data Quality - Process ] .flex-box[ ### Evaluate - Effectiveness - Outcomes - Equity ] .flex-box[ ### Improve - Feedback - Adjustments - Sustainability ] .flex-box.different-color[ ### Milestones - Evaluation Results - Improvement Recommendations - Sustainability Plan ] ]