Choose the easiest type of data source

The type of data source, such as existing sources or new instruments or systems, affects not only the time and resources needed to obtain and use data, but also impacts data validity and completeness and resources needed to manage and use the information.   Existing sources will generally provide the easiest way to get data, followed by compilation from existing sources, modification to an existing collection instrument, and finally, development and use of a new instrument.  Considerations for each of these types of data sources are identified below.

Existing sources

  • Readily available, doesn’t need a lot of maintenance to use (sometimes)
  • Ease of use will depend in part on if access to the data is automatic or requires ongoing effort (e.g. need to ask for data each time)
  • Need to understand source and data definitions very well

Compilation from existing sources

  • Readily available but takes some care and feeding to process
  • Amount of effort to compile will depend on data type (e.g. quantitative, categorical, detailed qualitative) and if any automation possible
  • Ease of use will depend in part on if access to the data is automatic or requires ongoing effort (e.g. need to ask for data each time)
  • Need to understand source and data definitions very well

Modification of existing instrument

  • Reporters are used to reporting using existing instruments; avoids startup data challenges such as extensive training needs, awareness, non-responses. Training on new fields will be needed.
  • Can tailor new data fields to your information needs
  • Start-up time and resources needed to train/explain/validate
  • Some resources needed to learn to work with the data as it actually comes in—might not be as expected

New instrument

  • Can tailor to your needs
  • Approval process may take time
  • Considerable start-up time and resources needed to train/explain/raise awareness
  • Considerable ongoing resources targeted to training and validation needed (reminding reporters how to report; reviewing data; making corrections)
  • Considerable ongoing resources targeted to nudging (reminding reporters to complete) also needed especially if there is no clear benefit to reporter
  • Some resources needed to learn to work with the data as it actually comes in—might not be as expected