Data quality rating (DQR)

Data quality rating turns a vague sense that “this dataset is not ideal” into a documented score that reviewers and users can interpret.

Criteria

ISO 14044 requires that data quality be described in terms of time-related, geographical and technological coverage, precision, completeness, representativeness, consistency, reproducibility and the sources of the data. The PEF method formalises this into a Data Quality Rating with scores for technological, geographical and time representativeness and precision, aggregated into an overall rating.

Pedigree matrix

Life cycle inventory databases such as ecoinvent use a pedigree matrix that scores reliability, completeness and the three types of representativeness, and translate the scores into uncertainty estimates.

Why it matters

A footprint built on outdated or geographically mismatched data may be numerically precise and still misleading. Data quality scores help users decide how much weight a result can bear, and PCRs and PEFCRs increasingly set minimum quality thresholds.

Where Verdatir fits

Verdatir captures data quality attributes for every input, checks them against the thresholds in the applicable rules, and surfaces low-quality inputs to reviewers before they propagate into a published result.