The government’s Data Service has today, Thursday 13 December,  published inaccurate Qualification Success Rate (QSR) reports for the second year running.

Providers use the reports to support the interpretation of their success rate results. Several colleges were quick to spot problems after the reports were made available for providers to download this morning, forcing the Data Service to issue an update on its website just hours later.

This follows similar problems last year when the QSR data had to be republished after qualification units were incorrectly included.

Our investigation into this issue has determined that there is an error and we have therefore removed the reports temporarily.”

A member of a forum for college data staff emailed members: “Ours are way out – it’s showing over 10,000 more starts! Also comparing last year’s QSR the 10-11 numbers are different by similar amounts.”

At about 4pm today the Data Service posted: “Following our earlier announcement we have identified a known issue with the sub-contractor reports in the 2011/12 Classroom Based Learning Qualification Success Rate reports published today. We are currently working on a revision to these reports and will shortly advise on arrangements for publication of this revision. We apologise for any inconvenience caused.”

The sub-contractor reports, which have already been identified as causing the problem, are new. The guidance notes for the QSR report states: “New subcontractor reports have been added to the report set. These show the success rate of the non-subcontracted and subcontracted provision, and the success rate of the subcontractors used by the provider.”

Jerry White,  director of planning and performance at City College Norwich, told FE Week: “It appears that the classroom QSR reports published today contain inaccuracies . . . If that is confirmed to be the case, then it is very disappointing that these important national data sets will need to be withdrawn and reissued for the second year in a row.”

More recently the National Success Rate Tables (NSRT), publicly available information about provider performance also produced by the Data Service, had to be temporarily withdrawn.

In June the Skills Funding Agency (SFA), on behalf of the Data Service, told FE Week: “The NSRTs had an unusually high number of unknown notional levels in the ‘other’ qualification type category.

“Our investigation into this issue has determined that there is an error and we have therefore removed the reports temporarily.”

As FE Week went to press, a spokesperson for the Association of Colleges said: “The AoC has not, as yet, had direct complaints or concerns raised with us from any of our members.”