How effective is score-based data quality assessment? An illustration with fish BCF data.

Area of Science:

• Environmental Science
• Ecotoxicology
• Data Science

Background:

• Rigorous data quality (DQ) evaluations are increasingly applied to environmental and ecotoxicological datasets.
• DQ is typically assessed by scoring data against predefined criteria.

Purpose of the Study:

• To examine the effectiveness of score-based DQ evaluations in statistically differentiating measurements.
• To use fish bioconcentration factor (BCF) datasets as a case study to assess DQ impact.

Main Methods:

• Inspected how log BCF differs based on overall and specific DQ evaluations.
• Analyzed interactive effects and hierarchy of DQ criteria on log BCF.
• Utilized tree analysis to understand deviations in log BCF based on criteria violations.

Main Results:

• 80-90% of chemicals showed no statistical difference in log BCF between low-quality (LQ) and high-quality (HQ) measurements.
• Log BCF did not consistently change with different combinations or numbers of DQ criteria violations.
• Simple averaging of all measurements yielded comparable log BCFs to HQ data (≤0.5 log unit difference in >93% of chemicals).

Conclusions:

• Data quality appears less critical for accurate log BCF than the number of independent measurements.
• Called for re-documentation of experimental details in legacy datasets.
• Recommended examining other DQ-categorized datasets and reflecting on DQ assessment methods for modeling and analysis.