mosaicQA - A General Approach to Facilitate Basic Data Quality Assurance for Epidemiological Research

Area of Science:

• Epidemiology
• Data Science
• Bioinformatics

Background:

• Epidemiological studies rely on extensive personal, medical, and socio-economic data.
• Ensuring high-quality data is crucial for reliable research findings.
• Continuous data quality assurance during collection is essential.

Purpose of the Study:

• Develop the mosaicQA-library, a set of R functions for non-statistical experts.
• Provide support for basic data quality assurance in epidemiological research.
• Accommodate a wide range of application scenarios.

Main Methods:

• Identified general quality-related questions.
• Designed R-scripts for metric and categorical data quality reports.
• Generalized R-scripts for flexibility and evaluated using diverse datasets.

Main Results:

• The mosaicQA-library generates basic data quality reports for multivariate data.
• Detailed single-variable reports can be produced, including definitions and missing value categories.
• The library is released as an official R-package for simplified installation.

Conclusions:

• Enables researchers to generate data quality reports for metric and categorical data without scripting knowledge.
• Focuses on data structure quality, assessing frequency, distribution, plausibility, missing, and extreme values.
• Supports data quality assurance in epidemiological research.