ISEB Atom dilemma - don’t understand

That maths question - it could come up in Stage 2 superselective maths grammar papers.

We did Atom this summer before grammar tests. The more DS practised the daily islands, the higher his score got. He did get full marks 142 in some mocks close to the exams and he did get good grammar results (superselective).

As far as I understand it from a quick google, Atom is not the official prep for ISEB anyway. It is Century Tech Bond Online this year?

Atom has taken some children that I know with no tutoring to excellent grammar results though in a short space of time and they offer a free programme to FSM children. So we should not diss them, they are doing their bit.

Best way to solve this is logic rather than common denominators. Compare to a half - D and E are both way less than a half so that rules them out,

1/4 is 1/4 less than a half
3/7 is half a seventh less than a half
2/5 is half a fifth less than a half

as 1/4 is bigger than 1/5 and 1/7 it must be smaller than both these. So it’s between 3/7 and 2/5. As 1/7 is smaller than 1/5, 3/7 is closer to a half than 2/5 so is the biggest.

looks complicated written down but IME kids who are very good at maths sort of work this sort of thing out for themselves. I’d say it’s beyond what I’d expect of most y5s though.

We didn't need to do adaptive tests for our area (sounds hideous) but I just ignored the sas scores and comparisons to others applying on atom and used it to practice the exam skills of completing a test on time and to cover the concepts that hadn't been done at school.
It could be argued it is in the interest of a website like that to under predict performance a) to avoid complaints if child doesn't pass despite doing well in atom tests and b) to worry parents that they need to keep using the website to do more practice... Their comparisons to other children applying to the same school are never going to be accurate regardless of their processes because not all applicants will be subscribed to atom.

Atom SAS are nonsense in my opinion as a teacher at a prep school. We used it to prep kids mostly for a state 11+. We did a mock VR and NVR test with about 12 kids and not one got anywhere near a pass mark 2 months before the exam. I raised this with atom who said they must have had a bad day. A whole cohort, twice?!? They pointed out one kid had got like 45% in the maths, which would not be a pass. Fair enough, except, as above, the questions were far far far too hard. That one above is definitely not y5 level. My top y5s could do it but it’s not y5 level. 2 months later, 7/10 kids who took the exam passed and the other 3 were very close - no one bombed like the atom scores suggested they would.

Atom is good for practice and homework but take all scores with a huge pinch of salt.

Agree there is a problem with how Atom is scoring. Very shady that they're attempting a cover-up.

This question, meanwhile, is from a mock test on Bond Online. Given the ISEB only goes up to the end of the Year 5 national curriculum in maths, this must surely be beyond the remit of the test?! There's no common denominator to be found here, unless they're expecting 11-year-olds to put all of these fractions over 16,380? Am I missing something?

@HanSB What was the technical fault? Did it have to do with the SAS scores or the adaptive tests?

There's definitely a technical fault with Atom, the IT team admitted it over email and I closed my account as they said that it couldn't be fixed. When I raised it in the whatsapp group and numerous others said they had the same problems, they closed the group down.

@Himawarigirl congratulations to your DC that's amazing. It is very interesting to hear about the iseb and atom. My son is getting quite low marks on the mock iseb tests and it has been quite a downer. He has his iseb test next week and I am really panicking!! I'm praying that he does better than he does on atom or by some miracle he gets the same questions he's been practising.

We are using Atom but also working with a very experienced tutor. She’s told us she thinks something is “off” with the Atom questions being disproportionally difficult - she’s finding this with a number of her students. Presumably this in turn could be skewing the algorithms and the SAS and comparative scores, I don’t know. She’s quite cross about it!

Unfortunately for us, our target school uses Atom for its entrance exam, so I can’t take any comfort in the knowledge it will be easier on the day! 😱 That said, if DC finds it hard presumably everyone will, and the scores will reflect that.

But for those discussing technique for these adaptive tests then agreed, the focus should be on taking time to get the questions right- especially the first 10 - rather than trying to finish them all.

That is the impression I got. Parents were questioning it on the Atom groups and they such the chat option so people couldn’t discuss it. Of course they used another excuse. It happened in multiple groups

Do you mean Atom are trying to keep their dodgy algorithm quiet?

Yes, we have found the same. Something is off and others have noticed it. Atom had whatsap groups closed because of it.

I don't know. And because the test is adaptive, it's hard to tell. But I agree with @Wlondon14 , something is not quite right about the SAS on Atom.

I don’t think the exam is harder, but the way the sas is presented might be inaccurate. Possibly, some children are taking so many practice exams they have had previous questions? or a minority with high scores taking hundreds of exams to confuse the sas grading?

Do we think mock ISEB maths tests on Atom are harder than the real ISEB maths test?

ISEB never transparent to student or parents their system either - it is also a black box.

Hi OP, on exactly the same boat (aside from DC with lower CAT score 😂). DC has always scored above average but when doing ISEB mocks they are suddenly average or well below average. I've complained to Atom. They just reply with some standard Oh, it's their first mocks, it's difficult, they need to get used to it and there's the adaptibility etc. Clearly, something is very odd in the scoring and they're not being transparent about it.
I have no advice. I'm just now glad that we registered for a mixture of ISEB and traditional paper tests schools and que sera sera. But I'm not happy. Atom should be transparent about their system.

I would ignore the Atom comparisons compared to average applicant. DS took the 11+ last year and was showing as average for the schools we applied for. Two of the schools we applied for gave feedback he was in the bottom 20% of applicants for exam performance so Atom was way off the mark. He scored top 2% in the country for his Year 6 SATS.

Go by test accuracy and use both paper and computer based. If your child is scoring well enough then that's all you need to know. We saw an exam consultant who gave the following benchmark for the London schools:

• top private schools (600+ applicants for 100 spaces) - score 95% in exam
• mid tier schools (500 applicants for 100 spaces) - score 80% in exam

I also can’t understand how their algorithm works, kagurazakamum. DD is also exceeding the average level of applicants to the most academic school we’re applying for, and to be honest, this is what I would expect, from the feedback we get from her school. (I appreciate of course that this is the level of the applicants, not those who were accepted.) But her Atom mock test results are still veering wildly between 85 and 142, with most of them in the 115-120 range, and I literally have no idea where they will land on the day.

I must confess that I have wondered a bit about the issue of test questions being repeated. We’ve had lots of issues with the same questions being repeated on Atom, sometimes multiple times in a week. It has occurred to me that if some kids are sitting lots of mock tests and getting questions to which they already know the answers (eg because they saw them yesterday), then does that perhaps drive the SAS curve up a bit..?

I am telling myself that it’s just an app and doesn’t actually have any bearing on her real schoolwork beyond this exam. Once she’s settled in a new school, her actual work will carry on as it always has done.

I have a similar dilemma here, for the atom mock exam results are quite shocking and never reflect the standardised scores DC normally gets from school reports. I appreciate DC might need to work further on the adaptive nature of some tests.

What does everyone else’s Performance Comparison’ look like on the other hand? We have put a top school as the target for now and DC looks to be exceeding in all subjects, which makes me doubt too.. They are compared against the applicants to the school at the time of their exam which kind of explains it, and also means we should look at it with good caution. Would that be right?

I would say the CAT test taken in Year 4/5 is quite different from the Atom/ISEB adaptive tests or other non-adaptive 11+ related online tests. The data sample for the latter includes students who have: 1. Undergone specific 11+ training and tutoring on all subjects, 2. Studied a more advanced curriculum for 11+ and 3. Taken the test multiple times to boost their scores, which can distort the normalized score. However, if your DS's mock performance isn't as expected across multiple platforms, perhaps consider applying to more schools. Unfortunately, for ISEB, it is a one-shot deal, so some dedicated focus on exam technique is required. Good luck.

I don’t know the answer, OP, but we are the same. CAT score of 140, but just can’t seem to click with the way the adaptive tests work. Sending lots of sympathy.

Although I do also think that some of the things they ask 10 year olds at a very early stage in the tests are absolutely insane - especially the vocabulary. We’ve had words in the first few questions that I honestly think a fair chunk of adults wouldn’t know. Last week, “omniscient” popped up at q.4, for example…I remember learning that in my A level history class! It’s hard to get the first few questions right under those circumstances…

My DD is going to have to do it all in one sitting as well as the school is not a centre, can only do at British council

@Kitkat189 Don't read too much into Atom scores. We didn't find them a useful indicator in the slightest. In our experience, they were very erratic, and bore very little relation to a child's actual abilities or the outcome of the 11+. Your son's CAT scores demonstrate he's naturally bright, so I wouldn't pull him out of top school assessments just based on practice tests so far. Have you discussed it with his teachers?