About Data Quality Statements

What is a data quality statement (DQS)? 

Every dataset in SEED includes a data quality statement. The DQS helps a user understand how a dataset can be used or compared with other, similar datasets. It includes characteristics of the data to help you decide whether the data will be fit for your specific purpose. 

Data quality statements (DQS) must be provided to SEED by the agency, custodian or manager of each dataset.   

Need to publish your data quality statements to SEED? Visit How to publish metadata and data quality statements.

What makes ‘quality’ SEED data? 

At SEED, we use a reporting questionnaire developed in line with NSW Government Standards for Data Quality Reporting. 

Data meets sufficient ‘quality’, or is ‘fit-for-purpose', when it meets its intended use (or re-use) in operations, decision making or planning.  

SEED’s data quality statement questionnaire uses a star rating system to rate a dataset’s strengths and weaknesses against key criteria. 

To get a star for a particular criterion, a dataset must achieve at least 4 out of 5 points.    

Be aware that the data quality rating needs to be interpreted with some care.   

• A dataset rated ‘Good’ for every data quality criterion may have zero stars.  
• A dataset rated ‘Poor’ for two criteria and ‘Very Good’ for three criteria may have 3 stars. 

While the data quality rating gives a quick indication of a dataset’s strengths we recommend referring to the Data Quality Statement for more detailed information. Users may make different assessments about the fitness of the same data, depending on the way they plan to use the data.   

How to prepare a data quality statement 

1. Use the reporting questionnaire to assess the data against five data quality criteria   
2. Institutional Environment, accuracy, coherence, interpretability, accessibility  
3. For each question: “yes” = 1 point; “no” = 0 points.  
4. Dimensions with 4 or 5 points receive a star.  
5. The data quality rating for the dataset is displayed on the dataset page.