Quality Management

Data quality management is composed of quality assurance (QA) and quality control (QC). Quality assurance begins before data are collected and are procedures used to prevent errors from entering the data (e.g., using a mobile app for data collection that limits possible values that can be entered (pick-lists)). Quality control is the discovery and correction of errors in the data and generally occurs during or after data collection (e.g., detection of outliers, typographical error, a character datum where a numeric value is expected, using an incorrect species code, and etc.). Quality control should occur as soon as possible after collecting the data and before submitting data to the archive record or sharing. QA and QC procedures should be identified during the project planning phase in consultation with the program’s biometrician and/or data manager. Record quality management practices (QA and QC) in the mdEditor for all documented products.

Best Practices in Quality Assurance

• Use documented protocols and standard methods

• Use high-quality instrumentation and regularly check accuracy

• Provide consistent training

• Develop standardized data collection forms (data sheet templates or computer input with data validation formats)

Best Practices in Quality Control

• Inspect data values using summary functions (tabling unique values, calculating means and variances, etc.) or by applying complex analysis algorithms.

• In Excel files, use sort and filter functions to look for data anomalies or outliers.

• Visually inspect data using scatterplots, regressions, and histograms.