Benjamin Feighery

QAR 2021 issues ( Downloaded CSV version)


Good Morning,

Has anyone else tried to reconcile these to their internal data yet?  it looks like they are significantly out, i am seeing for one of my clients learners who were submitted in 1920 as leavers who had planned ends in 20/21 being marked as 2020.5 in the hybrid year end (meaning they are expected to be included in a future year) so are not being included in the figures.  The .CSV file matches the dashboard data so it could be throwing the dashboard out by almost 30% in this instance.  anyone else found an explanation for this before i start complaining tot he ESFA?  

P.s.if anyone from the ESFA is reading this, it is really difficult to reconcile back to MIS systems where the 1920 hybrid year learners are completely missing and makes my job a lot harder!


No one has replied to this post.

Martin West

Yes similar issues here with Apprenticeships where not achieved EPA are 2020.5 but those who achieved are correct in 2020 but all should be in Hybrid year 2020.


Yup, had both of these and submitted queries.


Also, if you had any Standard learners where the AchDate (erroneously) wasn't completed last year, even though they completely finished last year, they're 2020.5 and I'm worried that, if we drag the records into 20/21 to correct them, they'll end up with HEY 2020 because of Reported Year...

Steve Bowler

Yes I've also submitted a query earlier as hardly any of our AEB aims are on the E&T raw report (over 2,000 missing)...


Benjamin Feighery

Thanks for the Replies,

it sounds like the issue might be with the way that they have redacted the 1920 data out of the files so hopefully an easy fix for them (fingers crossed before the R12 version)

Steve Hewitt i have the same concerns regarding the achievement date issues, i've had some heated exchanges with the ESFA team at the end of 18/19 about issues like this where it will end up double counting learner's across year but they don't seem to recognise the impact that has on the credibility of the stats or a providers ability to track their own performance. 

I still can't believe they are OK with forcing BIL's in to withdrawals if they don't return in the 2nd year and perfectly happy to count them again as withdrawals or achievers when they actually return; makes a mockery of the only limit to the length of a BIL being the qualification having closed before they decide to return.



Ben, the thing is bad people were BILing learners and never bringing them back, so they had to do something and this is a fairly reasonable compromise? Annoying that someone who goes on a BIL in July gets 11 months less than someone who goes in August, but that's the nature of the beast...

Benjamin Feighery

Steve Hewitt i think this response was disproportionate given they started badgering providers through the FRM's to look at their 12month + BILS at the same time, it was perfectly within their remit to write to providers with any volumes of long BILs and ask them to justify the situation rather than penalise learners; if they are going to impose something like this then they really have to review the impact on the rest of the rules and state the maximum duration of a BIL in the funding rules/provider support manual rather than buried in the QAR business rules.  



Funding Rules and QAR are two different things though (and they should be). As you say, when they come back they stop being a fail. I don't really think that's a problem (he says, being a little provocative...)?


In other news, in my E&T data, I've got 250 withdrawals, all QAR Inclusions, all past 14/42 days as appropriate, all planned (and actual) end dates before 31/7/21, all HEY 2020.5...

Benjamin Feighery

Steveh not so much a problem if they come back and achieve.. across years it's zero sum and a disappointment, but if they come back then withdraw or fail then you are hit twice for the same learner.  when we had the minimum standards in place, that one learner could be the difference between 40% and a visit from your contract manager.

counter provocative statement: ever feel like they might want to let a mistake like this through so that the QAR shows the rates are rising and claim that policies are working?

Martin West

One of the requirements for a BIL is that the reason for the break and its expected duration must be agreed with the employer and that there is an intention to return to the same apprenticeship programme but how many providers follow up on this and withdraw the learner when they do not return when agreed.

I think the QAR is ok as it is but that some providers want to game the system.

Emma Ford

Morning all,

Does anybody know if there have been any further developments on this?  Our dashboard is still visible and still contains the exact same information it did on the day of release.  Our Microsoft Excel data extract also still incorrectly matches this??

Many thanks


Benjamin Feighery

I just got a reply from the service desk to say the issue would be resolved in the R12 refresh so sounds like we can ignore this data and wait... 


Well that's a lot of flipping use, isn't it?

What if it's still wrong then???



We are aware that there is an issue with Hybrid End Year where Ach Date is blank (because the aim has withdrawn) and these are currently being marked as Hybrid End Year ‘2020.5’ when they should be marked as the Actual/Planned/Report Year depending on the latest of those three dates. This will be addressed and resolved in the next release of QARs In_Year at R12.

(this was specifically about App data, but I've got a horrible feeling it's the problem with my E&T aims as well which would be even more ridiculous...)

(ETA: yup, it's the problem with the E&T data too, what the hell was achdate doing anywhere near the E&T calc???)


Chris Bennett

Thanks for sharing that Steveh - exactly the same problem here with App data, number of achieved learners is as expected, but no withdrawals have gone into Hybrid End Year 2020, they are all 2020.5

It's a shame they cannot address, resolve and re-release the R10 data, or do an additional QAR In Year release at R11 given this problem. This would allow providers some extra time to check/reconcile against their own records and raise queries if needed.