HE institutions make data returns to many different bodies and there are many cases where data about the same students are being submitted in multiple data collections. This results in a need for the data collectors to understand and reconcile differences between these different data collections. In some cases this work is done within a data collection process as part of the queries that are raised with the institution.

Analysis of return specifications and submissions

Each data collection has its own unique data specification. Although there are some areas where data definitions are standardised the majority are not and the mapping from institutions internal data to the external data specification (as described in stage 1 of the model) is a bespoke process for each return at each institution. Therefore, queries resulting from the reconciliation between two different returns require the institution to undertake a comparative anlaysis of this mapping for these two returns in order to establish how their data manifests itself in each of the returns, in the context of the queries that are being raised.

Differences usually fall into one of three categories:

  • A data error in one or both of the returns
  • A mapping error in one or both of the returns
  • A legitimate difference in the mapping for the two the returns

The burden associated with understanding the external data specifications is broadly fixed. The burden associated with performing the analysis between the different datasets will vary according to the fit between the internal and external datasets and the nature of the differences that are being investigated.

Explanation and/or remedial actions

Having established the cause of the discrepancy the institution will enter into a dialogue with the data collector/collectors to explain the differences and agree what, if any, remedial actions should be undertaken. These discussions can be lengthly and complex and in some cases will result in the need for a previously signed-off collection to be resubmitted. In some instances a change is levied on the institution for this resubmission.

Back to the data burden projectNext: Funding and regulatory metrics