Graduate Outcomes statistical releases - webinar recordings
As part of our approach to the dissemination of Graduate Outcomes statistical outputs, we have delivered a series of webinars about the contents of each release. This page shares recordings of these webinars and provides answers to some of the detailed questions raised.
2019/2020 data and statistics
Live tour of the release starts at 4:30. Question and answer session starts at 1:02:32.
Questions and answers from the webinar:
All filters provided last year are all available on this year’s release so there will be a ‘provider type’ filter that allows filtering between HE and FE providers. We have also provided a csv mapping file which contains a list of providers by UKRPN, INSTID, name and provider type.
The main reason is the size of the tables – including additional level breakdowns will make them unresponsive and at their technical limit. We’re also aiming to meet user needs and we’ve not received this feedback before. Please let us know which tables you’d like to see this on and we might be able to do this for next year.
Yes this is available on Tables 1 to 4 which is higher education provider by activity, cross tabulated by personal characteristics and Table 6 which is provider by activity (and also includes proportions by confidence intervals).
Yes in the open data repository. You can use the search box to search by provider or Tables 1, 2, 3, 4, and 6 offer this but there also a number of others in the release that breakdown by provider.
All percentages in the release are calculated on the unrounded raw numbers. We calculate those percentages as the last step before we output that data and we round any values. Calculation on rounded numbers will provide errors. If there are specific tables you want to see percentages on we can consider that for future years.
Depends on the table. We typically use XACTIVITY, for example Table 8 looks at whether a graduate was running their own business, self-employed or developing a portfolio. We do use MIMPACT and ALLACT to split those specific ones out. We’ve released a table and charts specification which talks through all of the restrictions that we’ve put in place for the various tables and charts in this release so this will be helpful to users.
We paused work on a new performance indicator (PI) while we worked to figure out what we would produce in the space left by the discontinuation of UKPIs. We’re still in the process of going through those internal discussions about what data and measures would be most useful to our users and the sector. When we know more, we will provide more information. Find out more about our decision to cease production of UK Performance Indicators.
We’re moving away from producing data that would be branded ‘performance indicators’ so we need to be careful about avoiding data duplication in the sector. We would not want to produce a HESA version of the B3 metrics that are produced and released by the OfS.
We didn’t look at region as part of this research. We did consider this but once we had got through the number of activity categories and the number of government office regions, we were likely to be looking at very small pots of data in each of those cross tabulations, so we weren’t sure we’d be able to see much that was useful or publishable. The other reason is that in looking through the regional information about how the pandemic restrictions changed, they were often at a much more granular level than government office. This meant we weren’t entirely sure if we could get good data, would it even get to the granularity of which areas were under lockdown or which areas were feeling any hard, stark effects. Often differences are between urban and rural areas within a region and at the moment we don’t have geographic data that’s of a fine enough granularity to get at that. We are working to improve our geographic data so we’ll hopefully see an improvement in the future.
2018/19 Statistical Bulletin
Questions and answers from the webinar:
These can be found in the Metadata and definitions section of the Graduate Outcomes User Guide or you can download them directly from here: Tables and charts specification 2018/19
2017/18 Statistical Bulletin
Live tour of the Bulletin starts at 20:30. Question and answer session starts at 37:31.
Questions and answers from the webinar:
We include all graduates who either fully completed or partially completed the survey; these graduates have known outcomes. In addition, some tables include counts of non-respondents by personal or course characteristics. Some of the tables and charts within the release will relate to a specific restricted population but this will be clear from the titles on each table or chart.
The filters are labelled in a consistent way across the tables and charts and they will be consistent with concepts like mode of former study or level of qualification. We have also published definitions and expandable question boxes that explain more about what's included in particular filters.
In response to feedback, we have now published additional supporting information (Excel).
The Excel file specifies the data fields used within tables and charts in the releases. This includes the field names and restrictions applied to base tables from which the tables and charts were created and information about additional derivations we have applied to the data for the purpose of the data releases. Each base table references a list of tables and charts to which they relate.
We believe that the breadth and depth of material that we have published about Graduate Outcomes surpasses any previous release of data from HESA. We've really tried to be as open and transparent as possible about every aspect of the survey and how we present these statistics. But if users feel that there's further information needed, please get in touch with us at [email protected].
Yes, this is restricted to graduates with a most important activity of either paid work for an employer or voluntary work for an employer. The work type marker can filter between the most important activities.
The salary question in the survey is only asked of graduates who are in paid work for an employer or self-employment / freelancing, so it is just these that have been included in Figures 13 and 14.
There are currently no plans to look at international salaries due to issues around identifying relevant currencies and exchange rates.
There is validation within the survey itself and we ask graduates to double check their figures if they think they fall outside of what we would accept as the normal range. If a graduate confirms that it is correct, then they are free to enter that response and move on in the survey. We will look at where we can make improvements, where necessary, over time.
Further information on how confidence intervals are calculated is included in the footnotes to the tables that display confidence intervals. HESA uses the Goodman method.
We’ve included partial responses in our statistical outputs because they include usable survey data. We have no reason to believe that the quality of the responses is any less than the responses of graduates who have a completed response.
This depends on the particular measures in question. With partial responses, there will be high levels of unknowns in certain questions. If the measures in question are based on survey questions where we don’t have a full coverage of responses, then that will be taken into consideration in terms of how those are defined.
In addition to complete responses, partial responses are also SOC coded as long as we have usable data in the JOBTITLE and JOBDUTIES fields.
The work population marker filter displays two options: ‘work is an activity’ and ‘work is most important activity’. ‘Work is an activity’ is derived from all graduates who report an ALLACT 01-05. ‘Work is most important activity’ is further restricted to all graduates who report a value for MIMPACT within 01-05.
Our definitions page describes how we work out if somebody was in significant interim study or not. Essentially, we include any graduate who has completed a full-time higher education course or professional qualification between the points of graduating and the census week. On particular tables, we allow you to exclude that subset of people, depending on what you wish to do with the data. We have provided an expandable box within the release that provides this definition plus an example of why you might like to exclude these types of graduates.
At the moment, this is derived from the Graduate Outcomes survey itself. There is a section within the survey which asks graduates about any further study, training or research they were doing in the interim period. In the future we would like to start linking data sets to find students in our range of student records. This is not something that we’ve been able to do for this particular release.
This depends on the user’s purpose for the data. For example, a user may want to exclude those in significant interim study to assess the rates of unemployment among graduates. This is because among other examples, there will be some first degree graduates who go onto full-time master’s courses which they complete just before the census survey week begins. These graduates will have had less time to be looking for work compared to their peers who completed only the first degree. When looking at unemployment, it may be more suitable to exclude those who report to be in significant interim study.
We are using the JACS classification for the subject data within this release.
No, we are not planning on doing this during our first year of dissemination.
No, we will not due to data protection reasons. If users require unrounded figures, please get in touch with our colleagues at Jisc who compile bespoke datasets specific to your needs. There is a fee for this service.
The main vehicle for delivering data for competitor analysis will be via Heidi Plus which is delivered by Jisc. We are in discussions with Jisc about defining the set of appropriate Graduate Outcomes data to be released through Heidi Plus. This is a complex process and we are looking to focus on this work over the coming weeks.
This is not something that HESA can answer. It will be a decision for particularly the OfS and the other administrations in the UK who feed into that discussion. Organisations like the OfS will need to make assessments on quality and decide whether they feel that they fit for a particular purpose.
Our work with league table compilers has already commenced. We will be meeting with as many league table compilers as possible to ensure that they fully understand the data, what it’s capable of providing and that we understand what their expectations are and what they’re trying to achieve from the data.
In terms of the courses that are in scope for Graduate Outcomes, it’s not just degree courses. There may be HNC or HND type courses that are not degree courses. A full list of qualifications that are in scope for the survey can be found in the coverage page of the coding manual.