From Court
to Community:

It took several years, iterations, and rounds of feedback from multiple sources to create the measures. We consulted experts in the field of criminal justice research and measurement, including practitioners, researchers, and designers of measurement and information systems. We piloted the measures to see what worked and what didn't, and then passed the revised measures through another round of reviews by a second group of national experts in the field of criminal justice research and measurement. Finally, we sought feedback directly from practitioners in criminal justice (e.g., judges, prosecutors, police, public defenders, advocates, and community leaders) to ensure the measures are meaningful and accurate.

We are committed to publishing and sharing high-quality data. We conduct multiple rounds of internal and external reviews, and only data that meets our standards are published. In general, when we evaluate a dataset, our standards rely on the “Six Cs”:

1. Completeness: The degree to which a dataset doesn't include blank fields or otherwise missing data.
2. Correctness: The data is entered accurately and in accordance with the rules laid out in an agency's or court's data dictionary or codebook.
3. Consistency: The data is entered in the same way by each person and in accordance with the rules of the data dictionary.
4. Current: Data should be up-to-date and as close to real-time as possible.
5. Comprehensiveness: The degree to which a dataset paints the full picture by including data from all relevant parts of the system, such as police, prosecutors, courts, and more.
6. Contextualized: The data is evaluated in consideration of the data limitations, local reality, and statutory context of the jurisdiction it represents.

For a more detailed explanation of the process, please see the Data Quality Control section on our Methodology page.

We understand that every county is unique, and that counties across the country have different demographics, cultures, and practices that stem from those cultures. This is why we offer contextual information that enables practitioners to identify similarly situated counties, and to compare how similar cases may be handled in different counties.

Comparisons can be made because we:

• Standardize all data using common characteristics shared among disparate cases, which ensures we are, in fact, comparing apples to apples.
• Work continually to seek local input, local context, and information at multiple stages that is then passed on to users to help ensure data is interpreted responsibly, and enable the public to fully understand the limitations of the data.
• Make sure every measure is accompanied by key information to avoid possible misinterpretations of the data; this includes:
  • Footnotes. The footnotes provide technical information about individual state or county practices that may be germane to the measure.
  • Legal context. Each measure includes relevant statutes (and links) to explain specific differences in state policies that may affect criminal justice outcomes (e.g., state A uses sentencing guidelines while state B doesn't, etc.).
  • County context. We also include county context, which includes information on demographics, population, crime, criminal justice resources, poverty, commute times, drug hospitalizations, and citizenship. This allows users to compare counties that are similar in size, demographics, etc.

Rather than developing state- or county-specific data processing protocols, our measures are populated according to a standard methodology so that users of the national data portal can make consistent comparisons across jurisdictions.

Our methodology is rigorous, but is ultimately just one approach to data processing. Data can look different across sources because there are many valid ways to measure similar concepts. We include footnotes with measure calculation details and additional context so users can clearly identify where methods diverge.

We aim for these measures to be one tool communities across Michigan can use to better understand the criminal justice system, and encourage users of the National Data Portal to explore other sources in conjunction with our measures to see what new questions and insights they uncover.

With a couple of notable exceptions, all of Michigan's trial courts report case information to the statewide Judicial Data Warehouse (JDW) where data is then standardized to some extent. Importantly, the circuit and district courts in Berrien County and the 61st District Court of Grand Rapids did not yet report to the JDW at the time of extract. We collaborated with the Michigan State Court Administrative Office (SCAO) to ensure these exceptions are clearly noted for users of the data.

MI SCAO's efforts to standardize data from local courts laid a solid foundation for us to begin processing data for courts across the state. Still, varying approaches to data collection and recording at the local level lead to differences in the quality and completeness of the data included in the JDW and ultimately, the measures displayed here. Prior to publishing the data, MFJ met with the SCAO data experts to review these differences and to identify opportunities for further investigation at the local level.

Ideally, every state court system would have the data available for every measure. Because each county often has its own system to collect and record data, the data isn't always available or reliable enough to use in each county. This is why some states have more measures than others.

In 2022, the MI SCAO reached out to MFJ with a request: Help the courts improve their data, make it transparent, and engage Michiganders so that everyone can access and use the data. MFJ has worked closely with the SCAO to ensure the data accurately reflects how cases are processed in county courts across the state and has now shared that data with communities.

To start, join the MFJ Collective! It's full of information, webinars, and tutorials about how to make sense of data. It's also a great place to connect with other Michiganders who are using the data to inform their work.