Constructing a toolkit to evaluate quality of state and local administrative data

Area of Science:

• Public Administration
• Data Science
• Policy Analysis

Background:

• State and local agencies utilize administrative data for policy analysis and program evaluation.
• Effective use of administrative data requires understanding its quality.
• Agencies often lack resources and training for rigorous data quality assessments.

Purpose of the Study:

• To develop tools for assessing the quality of administrative data.
• To address challenges in constructing data quality assessment tools.
• To support agencies in improving their data for policy analysis and program evaluation.

Main Methods:

• Developed a toolkit focusing on critical data quality dimensions: accuracy, completeness, and comparability.
• Incorporated methods to evaluate longitudinal data components.
• Utilized data visualization to identify outliers and potential quality issues.

Main Results:

• The toolkit provides checks for accuracy, completeness, and comparability over time and across subgroups.
• Longitudinal data features are leveraged for enhanced data quality evaluation.
• Data visualization highlights records or variables needing attention due to quality concerns.

Conclusions:

• The developed toolkit offers generalizable solutions for common administrative data quality analyses.
• Customized analysis building upon the toolkit is beneficial for dataset-specific particularities.
• Improved data quality through these tools can enhance policy analysis and government program effectiveness.