Page cover

Data Quality 2

There's no such thing as perfect data - in fact, one person's idea of how a dataset could be improved may look very different from another person's.

However, issues, anomalies and outliers in the data may in fact be valuable to users.

Research suggests that data users look for errors and anomalies in the data can be seen as an entry point to sensemaking. This suggests that flagging and highlighting these 'strange things' to users so they become more quickly aware of them might be preferable to flattening out data by making it 'cleaner'. Such flags might point to unexpected values (e.g. outliers) or inconsistencies in formatting or standard ways of reporting. Wrestling with these strange things often serves as the entry point to a deeper engagement and understanding of the data, allowing users to question their assumptions and initial understanding of the data.

As long as people are aware of the limitations of their data and these limitations are clearly communicated (e.g., through documentation), it can be factored into the decisions being made based on that data.

Last updated