1522.0 - Quality Management of Statistical Outputs Produced From Administrative Data, Mar 2011  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 22/03/2011   
   Page tools: Print Print Page Print all pages in this productPrint All  


In order to achieve a particular outcome a data need (requirement) needs to be defined in order to make sure that the outcomes are met. When considering using administrative data for statistical purposes it is important to define what the data needs are before assessing potential administrative data collections to fulfil this requirement.

The ABS Data Quality Framework (cat. no.1520.0) can be useful in assisting the creation of a list of questions to establish a data need in line with a given objective for obtaining the data in the first place.

To describe a data need it is useful to create a list of questions beneath each of the seven dimensions contained in the ABS Data Quality Framework (Institutional Environment, Relevance, Timeliness, Accuracy, Coherence, Interpretability, Accessibility). This helps to ensure that all aspects of quality are considered when defining a data need. These questions can then be used to help assess potential administrative data sources for their fitness for purpose for the required objectives.

Some examples of questions that may be beneficial to consider before acquiring administrative data are:

  • Institutional Environment
      • From what type of organisation is it acceptable to obtain data?
      • Do we care how the data has been collected? i.e. direct collection or collected from a third party, face-to-face or online form.
  • Relevance
      • What population are we interested in?
      • What geographic level of detail do we require?
      • What concepts and data items do we need?
  • Timeliness
      • When do we need the data?
      • How recent do the data need to be in order to still be useful?
      • Will we want to get a time series of this information?
  • Accuracy
      • Do we have an acceptable level of quality that is required for the data we require?
      • Do we want to include data that has been imputed?
  • Coherence
      • Do we want to compare or combine several sources of data?
      • Do we need data that conforms to particular standards?
  • Interpretability
      • Do we want a copy of the questions asked?
      • Do we require a list of definitions for the concepts measured?
  • Accessibility
      • What level of detail do we need for this data?
      • What is our budget for purchasing data?
      • What processes do we expect around the acquisition of the data? e.g. secure electronic transfer

For more examples of questions that could be considered when defining a data need beneath each dimension of the ABS Data Quality Framework please see 'Defining A Data Need', on the National Statistical Services website (http://www.nss.gov.au/dataquality/PDFs/DQO_Needs.pdf).