Data Linking Toolkit: Step 5 – Analyze Data

Figure: Step 5: Analyze Data

Step 5 represents the ultimate benefit of the data linking work—partners being able to answer questions that they could not previously answer. There are many tools and resources about different types of analysis, but analysis itself is outside the scope of this toolkit. Instead, this toolkit focuses on the importance of developing an analysis plan and identifying those responsible for conducting the analysis. The table below displays the roles of team members potentially involved in Step 5 activities.

Figure: Step 5: Team Members

Activity 5a: Develop analysis plan

In Step 2 (Assess Partner Readiness), when data linking partners formalize their use case and data partnership, they may generally discuss the analysis plan. However, only after linking the data do the partners have a final list of data elements and better understand the challenges related to particular elements. Creating and finalizing the analysis plan now—after selecting data elements, establishing initial matching criteria, and developing business rules—allows those most involved to consider and document any limitations to the analysis, including issues related to data quality and completeness. This understanding translates into a better analysis plan.

In this activity, data linking partners agree to and finalize the analysis plan. The plan should include detailed information about the analysis approach and methods. An analysis plan is especially important for large, ongoing data linking projects because more people and data are involved over time. In addition, considering and planning for all audiences interested in the results, as well as what will be disseminated to those audiences, may influence the analysis plan.

The tool for Step 5 includes a template to assist with developing a data linking analysis and a dissemination plan to assist with detailing the analysis as well as the use of the results.

Activity 5b: Conduct analyses

The Data Linking Analysis and Dissemination Plan template describes the types of analyses to conduct with the linked data and how those data should be shared and used. Data linking partners may need to continually refine their analyses to answer the initial use case and any subsequent questions. The plan should be updated as analyses, dissemination, and/or use changes.

Both data linking partners, as well as subject matter experts, should review preliminary analyses to determine whether the results are reasonable and within expectations. If the results are inconclusive or do not pass face validity with subject matter experts, the partners may need to recheck the matching algorithm and linking processes for accuracy. (If one or more partners has an established Data Review Committee, they could participate in this activity.)

Published July 2022.