Page 11 of 11

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Wed Jul 15, 2015 10:01 am
by Papermover
Thank you so much Jenny.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Wed Jul 15, 2015 11:01 am
by Elizabeth
Although this is a small group, the implications of the results are important. I hope the comparisons will be duplicated in other schools.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Thu Jul 16, 2015 4:10 pm
by chew8
I've now been able to get a few more results, and a friend with statistical know-how has worked out that the correlation, at 0.88, is very high indeed (1.0 is a perfect correlation). Apparently the level of significance is also very high, and I think that means that the results are worth taking seriously in spite of the small size of the sample - perhaps kenm can confirm.

Jenny C.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Thu Jul 23, 2015 9:25 am
by kenm
Jenny C. wrote:...Apparently the level of significance is also very high, and I think that means that the results are worth taking seriously in spite of the small size of the sample - perhaps kenm can confirm.
I didn't notice this until yesterday. I shall have to learn some statistics, but it's all on the net somewhere.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Thu Jul 23, 2015 10:46 pm
by kenm
chew8 wrote:I've now been able to get a few more results, and a friend with statistical know-how has worked out that the correlation, at 0.88, is very high indeed (1.0 is a perfect correlation).

The value I get for Pearson's correlation coefficient using a formula in a spreadsheet is 0.894, which is pretty close.
Apparently the level of significance is also very high, and I think that means that the results are worth taking seriously in spite of the small size of the sample - perhaps kenm can confirm.
I found a table here to give an estimate of significance. 17 values and a correlation coefficient greater than 0.73 give a one-tailed probability of 0.0005. To the best of my understanding this means that your sample has less than one chance in 2000 of being from a population with zero correlation between the two variables. Unfortunately, I have mislaid the book that gives advice on when to use one-tailed and when two-tailed probabilities, but it is fair to say that the level of significance is, indeed, very high in either case.

Of course, it is not surprising to find that these two tests, both of which measure decoding, give correlated results. Do you have KS2 literacy scores for these children, or any other measure of comprehension of text? A correlation coefficient of high significance between these two variables would make a good argument for paying attention to the alphabetic code.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Fri Jul 24, 2015 7:41 am
by chew8
Many thanks. kenm.

By the time I sent the figures to my contact, I had been able to get the results of another four children, so the total number was 21, not 17, which probably accounts for the difference between 0.88 and 0.894. The word-reading test which I use does contain many words which are straightforward decoding-wise, but it also contains words such as 'one' and 'said', so it differs a bit from the phonics check.

I don't have KS2 results for these children as they have only just finished Year 4 and won't take the KS2 test until 2017. That could be an interesting year nationally, as it will involve the cohort which did the first Year 1 phonics check in 2012.

Jenny C.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Fri Jul 24, 2015 4:06 pm
by kenm
chew8 wrote:I don't have KS2 results for these children as they have only just finished Year 4 and won't take the KS2 test until 2017. That could be an interesting year nationally, as it will involve the cohort which did the first Year 1 phonics check in 2012.
What does the KS1 literacy test measure?

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Fri Jul 24, 2015 4:18 pm
by Elizabeth
Reading comprehension and writing composition. Someone else may give you more detail.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Sat Jul 25, 2015 11:31 am
by chew8
You can find sample materials for the 2016 KS1 tests here:

https://www.gov.uk/government/publicati ... structions

Unfortunately I don't normally get to know about the KS1 results of children who have done the phonics check. In fact, I don't routinely know phonics check scores either. I just happened to have been given them for the children I've mentioned in ths thread.

Jenny C.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Mon Jul 27, 2015 11:35 am
by volunteer
KS1 is all teacher assessed these days. It's all very confusing. As well as their judgement (from APP grids used throughout the year I presume) old reading tests had to be used. The reading test was a written comprehension - for the past few years old tests have been used again and again within schools and combined in with teacher's knowledge of the child (from APP I am guessing) to form a KS1 reading test level. It's not externally marked and not necessarily delivered under strict exam conditions, the children could have seen the tests at home already as they are available on the web and it is supposed to be combined in with teacher professional judgement on the child.

All very confusing. I think it will be similar under the new curriculum but there are some national performance descriptors currently geing developed to help teachers decide whether a child is below year 2 standard, approaching year 2 standard, at year 2 standard, above year 2 standard, or at mastery level. There was a consultation document last year but the final descriptors are not available until this Autumn.

Re: 'Light-touch, phonics-based check' for Y1 children

Posted: Mon Jul 27, 2015 12:10 pm
by chew8
I realise that KS1 assessment has man loopholes, but can only go on referring to my experience since I started helping voluntarily at a junior school in January 2001. My role has always been to hear the reading of the whole Year 3 cohort - about 90 children a year. In many of the years, I’ve also been able to find out the KS1 results of the children. In 90% or more of cases, these results have tallied well with my subjective impression of the children's reading and with the results of the Holborn reading test which the school has administered in the autumn term each year. I’ve posted on this in the past:


http://www.rrf.org.uk/messageforum/view ... orn#p32528

That was in 2009, but I’ve been able to make more checks since then and my impression hasn’t changed. My own view, therefore, is that loopholes or not, the KS1 results are a fairly reliable measure of children’s reading ability. The fact that there now also seems to be a strong correlation between phonics check scores and later performance on a standardised word-reading test is also encouraging.

Jenny C.