Hi Etienne, SMAB
Doesnt the paper on this link suggest that 117 in NFER is well above the average, assuming that it is standardised in the usual wayhttps://www.nfer.ac.uk/research/assessm ... scores.cfm
If I'm reading this correctly, it suggests that 117 would put DC in the 87 percentile for NFER (not for 11 plus)
Not a statistician, but would be interested to hear the view of others
We've both contributed to 47guest's thread, so you will have seen there that I wrote:
121 on its own probably wouldn't impress, even though it's round about the 92nd percentile on a national scale.
1. A nationally standardised test will reflect children of all abilities! The Bucks 11+, on the other hand, is standardised for the cohort. It's thought to be a very high standard, reflecting both the area and the fact that it's 'opt-in'. Even though something like 30% are successful, it would be wrong to equate this to the 70th percentile nationally. I read somewhere that the qualifying score might be roughly equivalent to the 92nd percentile on a national scale, but that's probably unofficial.
2. So what's wrong with a nationally standardised 121?
We have to keep in mind that we're talking about evidence that's going to convince a review or appeal panel.
121 might look very borderline to a panel - especially taking into account uncertainties (such as whether the tests really were done under strict exam conditions), and confidence intervals (agreed - the 'true score' could be higher or lower, but the aim of the exercise is to convince
When I was hearing appeals, my feeling - in so far as one can generalise - was that panels tended
(I put it no more strongly) to be looking for scores around the mid-120s. I recall one panel member who preferred to see 130!
When giving advice on the forum, I have consider what a panel might perceive to be a 'safe' score.
Someone on the forum once wrote that it's like seeking to adopt a child - you have more to prove than an 'ordinary' parent!