Can't wait to suggest these in the next subject leaders meeting
October 24, 2025 at 4:05 AM
Can't wait to suggest these in the next subject leaders meeting
That would be really useful! Thank you for the offer.
October 21, 2025 at 8:30 PM
That would be really useful! Thank you for the offer.
And are you aiming for a symmetrical set of GCSE results? I'm trying to work out whether that matters
October 21, 2025 at 8:20 PM
And are you aiming for a symmetrical set of GCSE results? I'm trying to work out whether that matters
Oh so when you say above or below what do you mean?
October 21, 2025 at 8:16 PM
Oh so when you say above or below what do you mean?
Mathematically speaking I think that's only true if each question was independent. If success on one predicts success on another I don't think so necessarily
October 21, 2025 at 8:12 PM
Mathematically speaking I think that's only true if each question was independent. If success on one predicts success on another I don't think so necessarily
Why did you decide that a symmetrical, unimodal distribution was the ideal? Wouldn't a skewed distribution be a better indication of good teaching? Especially if you're using a mastery approach. (Not criticising, I really want to understand!)
October 21, 2025 at 8:10 PM
Why did you decide that a symmetrical, unimodal distribution was the ideal? Wouldn't a skewed distribution be a better indication of good teaching? Especially if you're using a mastery approach. (Not criticising, I really want to understand!)
So if their percentile for CATS was 40th and now it's 60th you might say they're making above expected progress?
October 21, 2025 at 8:05 PM
So if their percentile for CATS was 40th and now it's 60th you might say they're making above expected progress?
I've always thought the skew in A level maths is weird. Other subjects aren't as skewed. Maybe if you remove FM students they're less skewed?
October 21, 2025 at 6:46 PM
I've always thought the skew in A level maths is weird. Other subjects aren't as skewed. Maybe if you remove FM students they're less skewed?
GCSE maths grades nationally are quite bell shaped. But it's a weird thing to aim for because if you want to increase your grade 9s you'd have to aim to increase your grade 1s too!
October 21, 2025 at 6:36 PM
GCSE maths grades nationally are quite bell shaped. But it's a weird thing to aim for because if you want to increase your grade 9s you'd have to aim to increase your grade 1s too!
Wouldn't any distribution identify that though?
October 21, 2025 at 6:35 PM
Wouldn't any distribution identify that though?
Oh good someone who's been through this! I would love to know how you think it's improved the usefulness of the data you have. Also you will always have half of your cohort below expected - is that useful?
October 21, 2025 at 6:34 PM
Oh good someone who's been through this! I would love to know how you think it's improved the usefulness of the data you have. Also you will always have half of your cohort below expected - is that useful?
This sounds sensible. The overlap means you can still compare well enough to decide if someone needs to move up or down a set.
October 21, 2025 at 6:31 PM
This sounds sensible. The overlap means you can still compare well enough to decide if someone needs to move up or down a set.
That would make some sense. But it would probably be a flawed expectation. There's no reason to believe results will naturally follow a normal distribution.
October 21, 2025 at 6:29 PM
That would make some sense. But it would probably be a flawed expectation. There's no reason to believe results will naturally follow a normal distribution.
It's just so hard when someone is using maths badly 😭😭😭
October 21, 2025 at 6:25 PM
It's just so hard when someone is using maths badly 😭😭😭
I agree. They didn't use to do the same test, we used to have three levels, but now they have to do the same one, which means that the bottom set has to do a test containing things they were never taught.
October 21, 2025 at 6:01 PM
I agree. They didn't use to do the same test, we used to have three levels, but now they have to do the same one, which means that the bottom set has to do a test containing things they were never taught.
Yeah, if anything we should be trying to match it to the distribution of our SATs results, or to the distribution we're aiming for in year 11.
October 21, 2025 at 5:55 PM
Yeah, if anything we should be trying to match it to the distribution of our SATs results, or to the distribution we're aiming for in year 11.
They definitely won't. Frustrating though because I'd rather refuse to do it on the grounds that it's pointless. I'm going on maternity in four weeks though so I'm trying not to care
October 21, 2025 at 5:05 PM
They definitely won't. Frustrating though because I'd rather refuse to do it on the grounds that it's pointless. I'm going on maternity in four weeks though so I'm trying not to care
And what I'm really concerned about is that we have whole year groups making less progress than they should, which will be hidden completely.
October 21, 2025 at 4:15 PM
And what I'm really concerned about is that we have whole year groups making less progress than they should, which will be hidden completely.
It does seem a bit like the tail wagging the dog
October 21, 2025 at 3:45 PM
It does seem a bit like the tail wagging the dog
I'm not sure if you can tell, but it was a non maths teacher who came up with this idea
October 21, 2025 at 3:42 PM
I'm not sure if you can tell, but it was a non maths teacher who came up with this idea
I think they just mean that rough shape. They said that our assessments are badly designed if they don't make that shape. They want to track percentiles as a way of monitoring progress (?)
October 21, 2025 at 3:39 PM
I think they just mean that rough shape. They said that our assessments are badly designed if they don't make that shape. They want to track percentiles as a way of monitoring progress (?)
Well I guess you look at your current distribution and you add easier or harder questions in? Or take them out? I mean the easiest way would be to leave the assessment as is and just use a transformation on the results, but that's cheating really
October 21, 2025 at 3:31 PM
Well I guess you look at your current distribution and you add easier or harder questions in? Or take them out? I mean the easiest way would be to leave the assessment as is and just use a transformation on the results, but that's cheating really