Lisa Danzig, Associate Director for Performance and Personnel Management at OMB, spoke at this morning's GAB Forum about the ways in which government agencies can use analytics to improve their performance, and some of the challenges associated with data-driven decision making.
Ms. Danzig identified five key types of analysis:
- Descriptive Analysis: Used to summarize data.
- Relational/exploratory Analysis: Used to identify patterns and relationships in data.
- Inferential/hyothesis-driven Analysis: Used to test of theories of change.
- Predictive Modeling: Used to forecast the event occurrences and trends.
- Randomized Studies: Used to quantify causal relationships.
The determination about which type of analysis to use depends on the particular research question or problem posed.
In addition to outlining the different types of analyses used in the public sector, Ms. Danzig noted three challenges to conducing quantitative research: (1) getting access to relevant data (and the data collection process used), (2) linking analyses to meaningful outcomes and (3) building up sufficient analytical capacity to conduct meaningful research. Overcoming these three challenges will allow government agenices to leverage the massive and increasing amount of data that's now available.
A recording of the event is available here.