AAF - Past planned end date

I mean, mine do, but that's because I've only got one learner more than 365 past planned end (and her actual end was only a week past 365...).

We are pushing ESFA for more specific information on how they are counting things precisely so we can share it so everyone knows what is being counted where and when.

It *should* be two groups of learner in PPED:

1. continuers who, at the given month's end, are more than 180 and 365 days past their planned end 

2. completers who, on their actual end date, were more than 180 and 365 days past their planned end

Does yours not add up if these are the two conditions?

Hi,

I have managed to match the data for both R06 and R07, for the current number of apprentices past their end date, the over 180 days, and over 365 days. However this has only been possible by using the previous period end date, so as per their example (e.g 31/03/22 for R08), I would have to substitute this with 28/02/22. 

It make no sense to me to use the ILR reference date from the Previous period, but it does give 6 matching answers, and not 6 non-matching answers. I have used dates in 2023, 31/01/2023, 28/02/2023, not 22 as their example states, (or have I completely misinterpreted what they were trying to say).

 

 

Nope, alongside internal reports I have also tried using the Apps Indicative Earnings Report to apply the two sets of conditions - still no match.

 

 

 

Oh, that's a very interesting finding Martin... Charlotte, does that help with you, if you use the month before?

(I think we can assume 22 was a typo...)

Sadly no, are you both using Apps Ind. Earnings?

Starting to rethink my goal of being able to perform an internal prediction every month before dashboard update :-s

 

I happened to use the Access export from FIS but, as I say, only having 1 it wasn't a problem!

I can't see why using the indicative would be a problem though? Rather difficult to write down exactly how I'd do it, basically filter for ZPROG001, look at completion =1 for planned ends more than 365/180 days ago, then filter completion for 2 and do actual end minus planned end and look for those over 365/180...

 

Yes Charlotte, I have been using the Apps Indicative Funding report and counting against the ZPROG001 aim, for all in-scope aims. Where a learner has multiple ZPROG001 records resulting from a funding change (Employer related) the original record is not counted. Multiple records created by BIL are counted, and thereby for part of the cohort against which the % is calculated. Counts are on completion status, 1,2, 3, & 6; in this case we are looking at only 1 & 2, and then making a comparison between the planned end date, and either the Learning actual end date for completers, or the ILR reference date for in learning. (using the previous period).

I am not happy with the results that I'm getting, so I am investigating potential alternative methods of counting. I have not been able to match the AAF results for BIL. If only the ESFA could provide a detailed specification of how they have arrived at their figures, and a drill down facility to see which learners, it would make it so much easier to sort out the discrepancies.

Well quite Martin. If you've not seen me mention it before, when I questioned ESFA on this, they suggested there was little demand for such a document, to which I replied that they probably hadn't asked the right people then, had they? As I said above, I'm trying to get a bit more out of them (actually having a meeting in a couple of weeks) and will, obviously share it far and wide when I do.

If there’s “little demand” for an AAF tech spec, I dread to think to what extent they imagine providers want the QAR framework looking at. A “passing interest” maybe?

[apologies for derailing] QAR doesn't need looking at, we just need another measure that is more useful. QAR itself isn't intrinsically bad, the problem is people feeling it should be higher, and feelings and data don't mix ;)

Even if they don't release a document that explains in detail how they calculate each figure, it would be really useful if they provided the learner data behind the figures then we'd be able to match it up against our own data...