Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
589191 | Safety Science | 2014 | 13 Pages |
•A robust data quality assessment needs to consider internal data quality, external data quality, and the reporting level.•Current aviation safety reporting systems are insufficient.•The fundamentals on which aviation safety management systems are based upon are rather weak.•Safety levels cannot be compared across countries and organizations.•A change in the mindset of the aviation industry is strongly needed.
The quality of data analysis and modeling is dependent on its inputs and statistical analysis is of limited value with inappropriate data. This paper proposes a framework for assessing data quality using the example of airport surface safety, i.e. runway/taxiway safety. The nature of airport surface safety is such that there is a need to account for data from a number of stakeholders, who may possess databases differing in quality, and aggregate this data for subsequent analysis to provide robust safety assessment and mitigation. To address these issues, this paper proposes a framework for the validation of external data quality based on the underlying data collection and investigation processes. Multi-Criteria Decision Analysis (MCDA) using a linear model is applied to derive quantitative weights for twelve safety databases based on the quality of the underlying organizational data collection and investigation processes. The model takes eleven criteria in relation to possible error sources during data gathering and pre-processing, organizational safety culture, data accessibility, and the consistency of the reporting system over time into account. These weights combined with an internal data quality validation and an indication of the reporting level of an organization can give a robust indication of the quality of a database. This method is recommended for use for data quality assessments in aviation safety.