Beta.gov

About Data And Evidence-Based Decisions Most public policies and practices have never been scientifically tested. Determinations about how we educate our children, deliver healthcare, rehabilitate incarcerated individuals, or house those without homes are not based on strong and rigorous empirical analyses. The traditional model of research typically involves funding challenges, long time lines, and topics that may have little practical utility. The BetaGov model of research offers an alternative approach. BetaGov encourages and assists agencies to use data to make important decisions. By using a scientific approach, your trials may significantly increase the pace of learning within your own agency by identifying what works and what doesn't. Developing In-House Research Agencies should be developing innovations based on topics and areas that are relevant to their mission, clients, staff, and culture. There are many different methods and designs for testing innovations; deciding what is best will consider the specifics of your agency and idea. The "gold standard" for research has long been recognized as the randomized controlled trial (RCT). This design allows for easy interpretation of results in order to improve efficiency and performance. BetaGov supports the use of RCTs in designing research, while also recognizing that some innovations may not be testable with an RCT design. We will always endeavor to turn your idea into an RCT, but when this isn't possible, we can use other designs such as a group randomization trial (GRT) or perhaps other designs that don’t require random assignment to condition. How It Works We help our partners identify problems and test solutions, or test promising ways to innovate and optimize positive outcomes. The process usually starts with only an idea, and a trial is developed through discussions with agency staff. The collaborative process looks like this: Invite staff to identify an area needing improvement or an idea for promising improvement Develop methodology to test the new program, process, or procedure that includes essential trial components (e.g., participants, measures, procedures, outcomes) and prepare written "trial description" Prepare measures/scales to gather data, trial scripts, consent forms, and other materials for the trial Ensure leadership approval and logistical support from others in the agency Provide introduction to research design and methods through "Pracademia" training of trial team Identify launch date and randomly assign participants to trial condition Implement intervention and control group procedures Collect and analyze trial data Prepare one-page trial summary "snapshot" to describe trial and outcomes