Raw Scores to Standardisation Mark

[HarryM]

Just thought it would be interesting to get an idea of all what your DC got in the form of Standardisation from raw score relative to month born from Bucks CC.
my DS got 70RS (S-120), 70RS (S-115) and is October Born
DS Best friend got 69RS (S-119), 71RS (S-119) and is Jan Born

Do you guys find this unbelievable???

It would be great if we managed to get a group of RS and SS from Bucks 11+ with DC's month born to plot predictive outcomes.

[Jules7]

I find it incredible too. Out of curiosity I requested DD 's raw scores yesterday so I will let you know when I get the info. She is October born and did pass so these will hopefully help plot a line for October.
I reckon on paper 1, 71 or 72 was required to pass for October but significantly more for paper 2. Maybe there were alot of October borns who did very well on paper 2?

[HarryM]

great DS has a twin, she got 126 and 141. I have requested raw scores and will post as soon as i get them.

[yoyo123]

standardisation is a statistical tool to iron out blips from tests which are easier or harder than previous years. It is a way of ensuring that the top 25% (or thereabouts will qualify. Norm referenced ( i.e standardised) results are used a lot to compare a child to the "average" in that cohort.

This may be of interest
http://www.elevenplusexams.co.uk/advice ... planation/" onclick="window.open(this.href);return false;

[Sally-Anne]

Hi Harry

I have taken the liberty of moving this thread to Bucks. Although your curiosity results from an appeal, I think the information will be of wider interest to Bucks parents, now and in future.

We also have a couple of "resident statisticians", particularly the estimable "Tree", who will undoubtedly help you to understand what has gone on.

It is important to remember that raw scores do not really have a big role to play at an appeal other than possibly on a very near miss of one raw mark: "It was one mark in the wrong box."

If we get a fair number of scores posted on here I will collate them into a table (not the easiest task on here!) for quick reference.

We do normally have a rule on Bucks that scores above 121 should not be posted because they are not relevant for school allocations. In this case we can definitely relax that rule because the information will be highly relevant to the discussion, but only if people are posting raw scores for information.

Sally-Anne

[Tree]

Hi everyone

As some of you know last year i took a load of raw and standardised scores and used some fairly simple mathematical modelling to produce a table of raw vs ss by birth month for last years tests please see this post for details:

http://www.elevenplusexams.co.uk/forum/ ... 2&start=10" onclick="window.open(this.href);return false;

I would be very keen to do the same for this years scores in order to get a good fit i need a minimum of 10 raw and ss scores across a range of birth months (for each test) to get a good fit. If people want to leave their scores in this post or pm me with them then i can make a table for this year. If we can get some scores for over 121 that would be great i didn't know you could get these.

once i have enough I'll try to run it and get a convergence.

on the face of harrym's scores it would seem that the second paper was alot easier it is difficult to say anything about age standardisation from these 2 results, which is interesting cos last year there was little difference between the 2 tests

all the best to you all

[sugarshaksk1]

hi tree i would like to help you but how does one go about getting raw scores.

on the face of harrym's scores it would seem that the second paper was alot easier it is difficult to say anything about age standardisation from these 2 results, which is interesting cos last year there was little difference between the 2 tests

[stressedparent]

Hi tree,
Nice to hear from you, if it helps my daughter is an October born and she has scored 131 in paper 1 and 141 in paper 2. From the raw data collected so far, I could see that paper 2 was a lot easier with high mean score among October born, what I find surprising is the huge shift in standardised score from one raw score difference. I will need one more October born raw score to evaluate the average raw score and standard deviation among October born.

[stressedparent]

With some data now available, using simultaneous equations and standardised score formula, we could approximately determine the standard deviation and average raw score of the 2010 October born cohort , I have arrrived at an average mean score of 65.4 with a standard deviation of 4.61. A low standard deviation indicates the data was very concentrated near the mean. A small shift of raw score will produce a high shift in standardised score within one or two of standard deviation from the mean"3.25 Ss for every 1 Rs" This explains that the october cohort were very well prepared and not much between them. A raw score of 70 will produce 115 standardised and a 78 rs will produce a 141. The formula you can use to find your raw score from standardised score as follows;
Rs=4.61(Ss-100)/15+65.4.this only applies to October born cohort only second bucks test paper.

[Jules7]

Wow what extra raw scores did you use to devise this?

I don't know if we will be given the raw scores as she did get >121 but I will let you know if so to help with this modelling business!

Re how to obtain raw scores, I just sent a letter to the address on her results paper and enclosed a cheque for £10 as mentioned in the guidelines somewhere. I did mention that I was requesting them under the freedom of information act.

I think I would like to raise a point about the narrow discrimination window. Surely it would be better to have a slightly harder test?