CATS scores in Y4 & afterwards & suitability for selectives

[Cranleigh]

Yes in Y5 - KS1 SATS results given but I have received nothing more recently in the way of test results. I am told these are available on request.

I have been advised it might be counter productive to confidence if we sit the pre-test and are not likely going to end up taking the exam (if we fail)? I guess I am also concerned they may pass and not be academically bright enough to thrive at the school. My gut tells me the school know what they are doing & unlikely to get through the pre-test if unsuitable.

If you go for it in these situations & children love the school what should your message be in advance?

I have been advised that CAT test score will give ball park suitability given school concerned will use a similar tool. DC comfortably in top 20 per cent of large cohort I'd say but not top 5 per cent sort of level. Good all rounder.

[southbucks3]

Go for the test, don't make a big deal of it. The test will probably be Saturday morning, so he will not have a chance to see much more than the school hall where the test is sat, they will invite him and you back for a proper look if he is successful. (Well that is the way our local and v good selective indie does it). He is only 9, so probably like my little man, still all believing, so get your best acting head on, and tell him he is doing a test just to show everyone how clever he is. Realistically he does not ever need to know the results, so I do not see how it can damage his confidence?

Top 20% in cats is good, but not amazing, he was presumably 8 years old and didn't even know what he was doing at the time so it is no biggie, certainly not a test to base his entire secondary education around. Like I said my summer baby ds got average 126 in his year 4 cats, but did some money sums for me today, and calculated the whole lot in fractions not decimals, eg. A quarter of the total cost of the items came to £9.1/2, and this sort of error is not unusual for him. Therefore I am sure it works both ways.

[Cranleigh]

Thank you - that really helped me.

[mystery]

Have you asked for the result of the CAT you say they did in year 4? Does that help you in any way?

If as a parent you like the school I would do the pre-test.

Maybe do a range of tests for a range of schools without your daughter getting too involved in really thinking about what life would be like at any of them?

With these entrance tests I think luck and effort plays a big part. If your daughter understands this she may not find that a "fail" knocks her confidence. Placing all your eggs in one basket with one entrance test only has to the be the scariest approach I would have thought.

Our posts crossed Southbucks.

I'm pretty sure I sat my prep school entrance test without ever visiting the school. You can call them all practices for something else.

[Etienne]

Etienne - how did you get the scores ? Thx

When advising on appeals, we sometimes get sent the evidence parents are thinking of submitting.
CAT results come with a one-page printout showing a range of information, including confidence intervals.

If the school withholds this information, parents could request it under the Data Protection Act.

Where do you test your dc?

Normally CATs are administered only by the school.
http://www.elevenplusexams.co.uk/appeal ... cation#b39" onclick="window.open(this.href);return false;

Yes the confidence intervals are important - but this is assuming everything was ideal and "standard conditions" on the day.

Isn't the point of confidence intervals that they take 'variables' into account? Anyone looking at how wide the confidence intervals are would surely conclude that a single set of results ought to be treated with a degree of caution?

• • "A pupil’s score is only an estimate of true ability ….." (GL Assessment / NFER)
    
    "It is important to appreciate that, however carefully educational tests are constructed, an element of error is likely to appear in the results they produce. For individual children, marks and scores should not be taken completely at their face value; they provide only an estimate of a pupil’s ability. This is …. not so clear when a numerical value is given; its accuracy and precision can easily be overestimated." (GL Assessment / NFER)

Also the confidence levels are quite big, rightly so, as my ds1 got full marks in one, great marks in another and awful in a third, he simply got wound up in tests sometimes, but managed to stay collected for his 11+.

That is the problem. On the other hand, if a child did three separate CATs (e.g. one in year 4, one in year 5, one in year 6), and the results were consistent, one might well conclude that they are likely to be reliable.

I should think that the test publishers will put a considerable amount of information into their various technical briefings about the test to warn schools not to use the results in such a hard and fast way.
Unfortunately some schools do just grab hold of one result and view it in a particular way with potentially life changing consequences. It must be useful to have a very high cat score to wave around at appeal, and it must be helpful if a child is identified early enough through cat testing as not achieving highly enough. But in general, one test on one day is not reliable enough on its own for decision making in terms of what to aim at with a child nor how they will do in a different test.

Agreed! - although I hope most appeal panels would have enough experience not to read too much into a single CAT score.

[mystery]

Hi Etienne, no I don't think that the confidence intervals do take into account all possible variables ---- interesting point to ponder though. Might take a look at some of my standardised test manuals (no not CATs!) and have another think about this when I have time.

I'm always thinking it is interesting that we don't have confidence intervals for 11plus results. That would confuse things wouldn't it?! I am assuming that the confidence interval on the new Kent test (much shorter than the old test) will be much wider than on the old much longer test.

Has anyone ever put in an FOI for the confidence intervals on the results?

[Etienne]

Hi Etienne, no I don't think that the confidence intervals do take into account all possible variables ---- interesting point to ponder though.

But let's suppose for the moment that they don't, aren't the confidence intervals already so wide that we - and NFER! - are agreed that marks and scores cannot be taken completely at their face value?

I'm always thinking it is interesting that we don't have confidence intervals for 11plus results. That would confuse things wouldn't it?!

I should think admission authorities would be terrified!

Has anyone ever put in an FOI for the confidence intervals on the results?

Not to my knowledge. Besides, if this information were kept by a commercial body rather than the LA, it wouldn't be held by a 'public body' as far as I can tell. A DPA request, relating to a specific child's results, might stand a better chance - but again, it might be information that the LA doesn't have (and doesn't want!).

[mystery]

Yes - most likely! Selection is even more complex than most people think and the fewer questions asked the happier everyone is. It does make the search for the "holy grail" of admissions tests look more and more futile though. I think the idea is that, in the same way that Mrs Thatcher kept the 75% happy by reducing the number of grammar schools, these days we hunt endlessly for a fair test so that it seems as though everyone has an equal chance - thus keeping 100% happy?