37 - Ongoing, Dynamic, and Interactive Evaluations: Practical Guidance for New Policy Standards

Abstract Information: The 2022 Educational Policy and Accreditation Standards (EPAS) of the Council on Social Work Education (CSWE) added five elements to its evaluation competency (9). Evaluations should be: ongoing, dynamic, interactive, anti-racist, and anti-oppressive. This skill-building workshop will highlight strategic and technical practices to achieve these goals. Strategic: How agencies become learning organizations- We will describe essential steps to move evaluations from stress-inducing interruptions to normative feedback for improved practice. Using anecdotal evidence from one child welfare agency’s PQI experience, we will address issues related to power dynamics, data literacy, cultural competency, ownership, and interpretation of findings, including the stakeholders needed at each step of the process to achieve these goals. Technical: How to build a data model for evaluation that is: accurate, actionable, adaptable, and affordable- Ongoing and dynamic evaluations require some level of automation. Significant insights can be gleaned from the CI/CD pipelines (continuous integration/continuous deployment) employed in software development and data engineering, particularly in creating data models that support an agency’s PQI strategy. We will examine four aspects of a modern data architecture that help achieve the EPAS competency 9 goals: • Semantic layer: A single source of truth ingesting data from an increasing number of data sources. • Automated ELT with built-in testing and alerts. • Self-service analytics: Create data products that can be consumed by practioners. • Separate production and development environments. Each aspect highlighted will include an introduction to industry-leading and relevant technologies and tools (e.g., Snowflake, BigQuery, dbt). Lastly, we will showcase specific evaluation techniques that address the goals of anti-racist and anti-oppressive. These will include built-in assessments and inferential statistics on referrals and admissions (e.g., racial equity frameworks) and outcome evaluations. We will end the workshop by inviting participants to sketch out a diagram of a data model that addresses these five goals. Participants can use any social sector they would like. We will provide a demonstration for child welfare.