Data assurance

The calculation of teaching funding is based on, and verified against, data supplied by providers in the Higher Education Students Early Statistics (HESES) aggregate statistical survey (formerly HESES or HEIFES), and against Higher Education Statistics Agency (HESA) data and Individualised Learner Record (ILR) submissions. We perform data reconciliation work between these returns.

General method

Every year we generate re-creations for all providers by applying our algorithms to HESA/ILR student data to produce derived fields. These derived fields are then aggregated to produce re-creations of the original HESES aggregate return.

We then produce summaries and comparisons of the main elements of each re-creation against the original aggregate return and present these in an Excel workbook.

The necessarily complex process of explaining and resolving differences between data sources places a burden on providers and the OfS. To ensure this burden is both manageable and appropriate, we employ thresholds based on funding and/or headcount differences to select which providers must respond to a data reconciliation exercise.

Action required

Where the comparison of a re-creation and an original aggregate return indicates discrepancies exceeding any of the annual selection criteria, we ask the provider to respond with an action plan that details how they will reconcile the two data sources.

Explanations for discrepancies

We expect the explanations that providers submit to explain discrepancies between the two data sources to fall into one or more of the following categories:

• Errors in HESA data

Where errors are found in HESA data we require providers to submit a revised, complete and valid HESA return directly to HESA, but only once these changes have been notified to us through an action plan, and we have approved the plan. Subsequently the provider will need to detail the amendments in an error summary file (ESF) which will also need to be approved by us before amendments to data are made. 

• Errors in ILR data

Where errors are found in ILR data we require providers to submit to us amendment files, detailing the corrections to be made to the ILR return, but only once these changes have been notified to us through an action plan, and we have approved the plan. Subsequently the provider will need to detail the amendments in an ESF which will also need to be approved by us before amendments to data are made.

• Errors or estimation discrepancies in original aggregate return

If we find, either through data reconciliations or audit that the original aggregate return does not reflect the outturn position for the year, and this is due to errors or estimation discrepancies, then the re-creation will supersede the original return. Consequently, it will generally not be necessary for providers to submit corrections to the original return.

• Problems of fit with the re-creation algorithms 

We do not normally expect that problems of fit with the re-creation algorithms will fully explain discrepancies that exceed the selection thresholds. However, where a problem of fit between our algorithms and the aggregate return definitions contribute to a discrepancy, an explanation will be required of where the problem occurs, and its impact, through the action plan. In addition, providers will need to provide a primary derived field override file to enable us to correct the problem of fit with our algorithms for the data affected.

Further process details

As well as the completion of an action plan, the exercise will also include a visit to the provider to discuss the systems and processes used to compile the data returns. We will also review a sample of students with the provider, discussing any potential data errors during the visit.

Following the visit and approval of the action plan, we may carry out further sampling of the data to test the correct reporting of some specific key fields. If from any of these processes it is evident that data amendments are required to the HESA or ILR data the provider will be required to make these. Any proposed amendments to data are reported to us through an ESF which needs our approval before amendments are made. Once approved, the provider will amend their HESA or ILR data. Once the changes to data match the details in the ESF, we will carry out testing on the amended data to ensure the appropriate amendments have been made, and that no further material systematic errors remain.

If providers do not provide satisfactory explanations for discrepancies, or do not respond within the given timescales, we may carry out further investigations, which might include an audit. 

Uses of the amended data including funding implications

At the end of the reconciliation exercise we ask providers to confirm that the re-creation reasonably reflects the outturn position for the year and the data is signed off by the accountable officer. 

If selection for the exercise was on the basis of a funding or funding AND headcount difference, the re-creation supersedes the original aggregate return and any resulting grant adjustments are made, subject to any appeals process that may apply and the availability of our funds. The OfS data amendment panel will decide on other uses of the amended data.

If selection for the exercise was on the basis of only a headcount difference, the OfS data amendment panel will decide on all uses of the amended data, which may include funding.