Potential Kendrick Cutoff

[Langtreemum]

How many took the test last year?

[Tinkers]

The two papers are likely to be of similar score as they are both a mix of all subjects, unlike when my DD took it with separate VR and NVR, (there was a 22 point difference in her two scores, and I know another girl with a 25 point difference).

However I'm lead to believe from the school website that they are still giving an average of two papers. This will make the stats not 'quite' straightforward, though it will be reasonably close.

776 took last year. 651 the year before.

Last year we used the calculator to establish that 117 or 118 was enough to be sure of a place (I can't remember which), however we knew that the places would probably get down to 120ish and guessed that 115 would be the cutoff, in the end it was 114.

141 is still potentially the max score as it will be the max score per paper. However you aren't not going to get 141 unless you get 141 in both papers. How does it work of you use 141 as the max instead?

[Langtreemum]

OK, 776 students and number 120-125 being at 114 implies a sd of 13.5-14 with mean of 100.

So I stand by my hypothesis of standardising each one to 115, but it gets lowered by averaging.

[Curious__George]

There are several pointers to why the sigma is likely to be 10 and not 15, with mean being 100.

- The empirical rule states that 1 sigma is 68.26%, 2 sigma is 95.44%, 3 sigma is 99.73%
- The cut off is 110 and not 115 & there lies the clue. It is normal to expect the cutoff to be 1 sigma.
- if the sd is 15, percentile of the known top score would only be 98.96% which is way off from a statistical point of view

It looks like there seems to be some blind belief in the so called kendrick calculator, when the data points otherwise.

[Curious__George]

The two papers are likely to be of similar score as they are both a mix of all subjects, unlike when my DD took it with separate VR and NVR, (there was a 22 point difference in her two scores, and I know another girl with a 25 point difference).

141 is still potentially the max score as it will be the max score per paper. However you aren't not going to get 141 unless you get 141 in both papers. How does it work of you use 141 as the max instead?

141 has nothing to do with the raw scores are as it is a standardised score - a numerical view of what is essentially a percentile. i.e. 141 represents the 99.999th percentile. In simple terms, if the total marks achievable in both tests are 150 & someone scored only 120 but everyone else scored below that, then that student would get a standardised score of 141.

[Langtreemum]

I take your point about the cutoff of 110 tallying in some sense with a sd of 10 - but they might not want to set a cut-off as high as that. If they cut-off at one sd then they would only be potentially letting in the top 16% (remember that if you look outside one sd of the mean there are 16% at the top and 16% at the bottom to give you your 68% left in the interval). That might not be enough - they might want to consider more than 16% - and in fact on an entry of 700ish they surely would?

And a sd of 10 means that last year's numbers of going down 120-125 at a score of 114 don't tally. You have to both lower the number they took and lower the cut-off score to make it work eg only going about 100 down the list and a cutoff score of 110-11, but we know that didn't happen!

[Langtreemum]

Anyway - it's basically all guesswork - and even if we did know the actual sd (not sure why they couldn't tell us) there would still be lots of guesswork in estimating how many are either outside the designated area or won't apply in the end anyway.

And my understanding is there is no need to be strategic, we should just put down our preferences, so our speculations dont change our decisions!

(my son has actually just taken the Reading test btw, but my hypothesis is the standardisation is the same. My daughter is already at Kendrick)

[Langtreemum]

Of course they may have changed the standardisation sd compared to last year, but as far as I can see we have no reason to think that.

You seem to be putting a lot of weight on the max score and assuming they aren't truncating reported scores at the top. I think they probably are.

[gsplace]

Our dd got 119.35 for Kendrick, so not sure if it will be enough. Are there any chances to get in Kendrick? Fingers crossed

[Tinkers]

I think you have a reasonable shot.