Skip to main content

Methodologies

Is it known how long the online phase will be open before commencing telephoning, or will the two methodologies run concurrently?

They will likely run concurrently, although this will be agreed through the implementation phase.

Will each provider receive a minimum number of calling hours, or will the amount of investment in calling be determined by the online response rate?

There will be a level of quality in responses across providers which will be targeted. Where low responses are received online, or the online response rates are of poor quality then these graduates may receive more calling time.

We have concern over using institutional emails only. After 15 month the drop off in use of those is likely to be high. Why can't personal emails be accepted?

We are not proposing to use institutional emails only. Our preference will be for non-institutional email addresses, for the reasons set out above.

We support the idea of ‘local communications’ and would like further information on this? We've found that emails from Heads of Department positively impact email response rate. Emails to international students in their native language is similar.

HESA will be developing communications materials which will be made available to providers to support centralised communications. We will publish more information about local communications, and good practice for keeping contact details up to date, in due course.

We wouldn’t want the removal of the postal survey to have a negative effect on disadvantaged or under-represented student groups. Is it the case that HESA plan to perform some bias testing to understand this better?

This issue was researched when removing the postal survey for the Longitudinal DLHE. The research showed that there was no impact on these groups.

For the 2014/15 DLHE survey (C14018), only 2% of responses were received via postal survey.

Why are you centralising?

We have consulted extensively with the sector, with data users and with key official organisations. We have developed a model of open centralisation which enables us to collect the robust and rich data we need in an efficient manner, while still allowing HE providers speedy access to that data and a significant stake in the survey's governance. See Dan Cook’s blog for further information.

In terms of providing evidence for HEP third party responses, what additional information will be required to be a valid response?

This will be set out in the operational materials which we will begin publishing in June.

You suggest more online responses would reduce cost, and that there will be a drive to increase online responses. How will you drive this increase?

Online response rates have been going up year on year. We will learn from good practice across the sector. In particular, we will learn from Longitudinal DLHE - which is receiving significantly higher online response rates for the most recent collection (C12019).

Online forms overall at this institution gain lower-level quality of information when compared to phone contact. Would the contractor be expected to match current contact methods to ensure quality of response?

HESA are moving towards more responses being sought online, as fits with increasing pick up of this methodology. However we will build in high levels of quality assurance to the online tool, to ensure the data collected this way is of as high quality as telephone.

Is there a deadline for choosing the optional questions?

Yes, there will be. This will be set out in the operational materials.

Are the current HESA data protection notices sufficient to cover the first cohort of current students?

We are reviewing our collection notices now and will be issuing further guidance in June.

We understand that the survey will still be used to gather salary data for graduates working outside the UK. Has any analysis been carried out to understand graduate numbers working outside the UK?

Yes, from analysis it has been found that approximately 5% of graduates are employed outside of the UK (2014/15 data) excluding unknowns.