Federal Committee on Statistical Methodology (FCSM) Equitable Data Toolkit

About the FCSM Equitable Data Toolkit

The Federal Committee on Statistical Methodology (FCSM) developed the FCSM Equitable Data Toolkit (Toolkit) to provide federal agencies with tools that support equity analyses with a focus on historically underserved populations. It is intended to support an increase in available data, improve the accuracy of analyses, and ensure ethical and secure data governance to improve the representation of underserved populations in federal data and analyses. The choice of which populations to examine in the Toolkit was influenced by long standing measurement challenges in the federal data system.

It is noted that the FCSM provides materials on LGBTQ+ outside of the Toolkit on the webpage Measuring Sexual Orientation and Gender Identity (SOGI) Research Group.

Historically underserved populations are often "hard to measure" in sample surveys and even administrative data systems. In some cases, the population group is small, and its members are difficult to locate and include in probability-based surveys. In addition, obtaining the information needed to properly classify individuals as members of the population group of interest can be difficult.

Measuring experiences and outcomes for these groups involve a wide range of methodological challenges, including:

• Creating questions that members of the population groups recognize (regardless of the relative size of the subpopulation).
• Devising accurate proxy reporting measures for administrative record systems.
• Developing terms, scales, and definitions that can be useful in variety of contexts and studies.
• Implementing data disclosure protections that still permit public-release of information
• Interpreting statistical results when placing them into a broader context.

The Toolkit provides information and resources for Statistical Officials, Chief Data Officers, agency staff, and practitioners seeking guidance on:

• Refining wording of survey items to operationalize concepts that facilitate identification or self-identification of characteristics in ways that are relevant for equity analysis.
• Overcoming sample-size issues that affect the precision of estimates for population groups that are small in number, geographically diffuse, and possibly difficult to identify or self-identify.
• Linking across data sources to build information including matching strategies and evaluation of linkage success.
• Addressing data-protection issues that spring from a survey participant being a member of a relatively small population group and one of relatively few such members included in the sample; these issues are especially salient when data are examined for subgroups defined by a combination of variables (race/ethnicity, geographic location, etc.) that can result in unacceptably small number of observations in the subgroup
• Issues of definition and measurement involving geography and persistent poverty.

The information and resources provided here are general. Those who use the Toolkit will need to assess the extent to which it is helpful to strive for consistency across agencies and their data collections and when it is helpful to develop an approach tailored for a specific purpose.